6 Essential Skills to Make A Great BI Analyst. Series 1 | KAIMA LM | Skillshare

6 Essential Skills to Make A Great BI Analyst. Series 1

KAIMA LM, BI/Data Scientist

Play Speed
  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x
19 Lessons (1h 25m)
    • 1. Introduction: What does this course cover?

      3:15
    • 2. Why are there so many Jagons in Business and Data Science?

      4:38
    • 3. Analysis Vs Analytics

      3:57
    • 4. An Overview of Data Science Task

      1:20
    • 5. Traditional Methods of Applying Data

      6:44
    • 6. Traditional Data Techniques

      8:07
    • 7. Traditional Data Real Life Examples

      1:35
    • 8. Big Data Techniques

      4:02
    • 9. Big Data Real Life Examples

      1:26
    • 10. Business Intelligence Techniques

      6:41
    • 11. Business Intelligence Real Life Examples

      1:49
    • 12. 7 Traditional Methods Techniques

      8:35
    • 13. Traditional Real Life Examples

      2:38
    • 14. Machine Learning Techniques

      6:46
    • 15. Types of Machine Learning

      8:27
    • 16. Machine Learning Real Life Examples

      2:15
    • 17. Programming Language and Software Employed in Data Science

      5:48
    • 18. Data Science Job Positions

      3:09
    • 19. Dispelling Common Misconceptions

      3:51

About This Class

Hi! Welcome to our Business Intelligence Analyst Course Series. This course entails the six essential skills you need to make a great BI analyst

We are excited to present you a course series that stands out.

This BI program is different than the rest of the materials available online.  

These are the precise technical skills recruiters are looking for when hiring BI Analysts. And today, you have the chance of acquiring an invaluable advantage to get ahead of other candidates. This course will be the secret to your success. And your success is our success, so let’s make it happen!  

  • Introduction to Data and Data Science  
  • Statistics and Excel  
  • Database theory  
  • SQL  
  • Tableau  
  • SQL + Tableau  

Here are some more details of what you get with The Business Intelligence Analyst Course:   

  • Introduction to Data and Data Science – Make sense of terms like business intelligence, traditional and big data, traditional statistical methods, machine learning, predictive analytics, supervised learning, unsupervised learning, reinforcement learning, and many more;  
  • Statistics and Excel – Understand statistical testing and build a solid foundation. Modern software packages and programming languages are automating most of these activities, but this part of the course gives you something more valuable – critical thinking abilities;  
  • Database theory – Before you start using SQL, it is highly beneficial to learn about the underlying database theory and acquire an understanding of why databases are created and how they can help us manage data  
  • SQL - when you can work with SQL, it means you don’t have to rely on others sending you data and executing queries for you. You can do that on your own. This allows you to be independent and dig deeper into the data to obtain the answers to questions that might improve the way your company does its business  
  • Tableau– one of the most powerful and intuitive data visualization tools available out there. Almost all large companies use such tools to enhance their BI capabilities. Tableau is the #1 best-in-class solution that helps you create powerful charts and dashboards  
  • Learning a programming language is meaningless without putting it to use. That’s why we integrate SQL and Tableau, and perform several real-life Business Intelligence tasks  

 Sounds Awesome, right?  Our course comes with

Our courses are unique and is equipped with:

  • Work with real-life examples  
  • Provide easy to understand and complete explanations  
  • Create beautiful and engaging animations  
  • Prepare exercises, course notes, quizzes, and other materials that will enhance your course taking experience  
  • Be there for you and provide support whenever necessary  

We love teaching and we are really excited about this journey. It will get your foot in the door of an exciting and rising profession. Don’t hesitate and subscribe today. The only regret you will have is that you didn’t find this course sooner!

What you’ll learn

  • Become an expert in Statistics, SQL, Tableau, and problem solving
  • Boost your resume with in-demand skills
  • Gather, organize, analyze and visualize data
  • Use data for improved business decision-making
  • Present information in the form of metrics, KPIs, reports, and dashboards
  • Perform quantitative and qualitative business analysis
  • Analyze current and historical data
  • Discover how to find trends, market conditions, and research competitor positioning
  • Understand the fundamentals of database theory
  • Use SQL to create, design, and manipulate SQL databases
  • Extract data from a database writing your own queries
  • Create powerful professional visualizations in Tableau
  • Combine SQL and Tableau to visualize data from the source
  • Solve real-world business analysis tasks in SQL and Tableau

Are there any course requirements or prerequisites?

  • No prior experience is required. We will start from the very basics
  • You’ll need to install MySQL, Tableau Public, and Anaconda. We will show you how to do it step by step
  • Microsoft Excel 2003, 2010, 2013, 2016, or 365

Who this course is for:

  • Beginners to programming and data science
  • Students eager to learn about job opportunities in the field of data science
  • Candidates willing to boost their resume by learning how to combine the knowledge of Statistics, SQL, and Tableau in a real-world working environment
  • SQL Programmers who want to develop business reasoning and apply their knowledge to the solution of various business tasks
  • People interested in a Business Intelligence Analyst career

Transcripts

1. Introduction: What does this course cover?: I and welcome to our business intelligence analyst course, Siri's. This course entails the six essential skills you need to make a great BR on a list. My name is Karima, another business intelligence analyst. Was spent a significant amount of time walking on business intelligence dust requiring more skills like statistics, Excel, database story, Bablu, beytin, the combination off SQL and tableau on even the combination of SQL with tableau and biter. I'm really excited to present to you because Siri's that stands out because each other call Siri stitches. You is done alone skill you need a Xabi are on a list, and by the end of critical Siri's, you would know how to apply Eat in the real world walking environment they cost. There is curriculum that has been prepared for you. Consists off diverse sets off topics are serious. Start with getting you to know the world of data and data signs. We explain all that to Science Jagan's on areas of activities before diving into more sophisticated analytical. Thus, once we have built a solid theoretical foundation, we go through statistics applied in excel and teach database management in s key. Well, then you will be ready to land disabilities off data of data visualization are reporting by creating professional visualizations with Bablu public, one of the most popular, be Iittle available all day. What I consider to be remembered if it's off this program, is that you will let out of leverage, ask your skill and combined them with Bablu to visualize the data contained in disobeys files. Once you know how to do that, it is time to start accordion infighter. But pretty soon he will understand that Lenin, a programming language, is meaningless without putting it to use. That's why, in final Siri's off, this course will integrate SQL bite and and taboo, which would allow us to build a model that predicts clients default rates on. Visualize your findings in taboo, and this will take your preparation to the next level. Because Siri's is it truly also adventure. To get the most out of this series journey, please don't skip any of the lessons in every Siri's as well gradually beauty or knowledge . Given that we teach several skills, it is essential he starts at the beginning as this lessons at the foundation, you need to tackle the more advanced topics you finally town in the program. Many of our lesson contained downloadable resources that will help reinforce what you've learned, such as cost notes, exercise files, pdf materials and the notebook files. Everything is included and can be downloaded easily. I strongly suggest you complete all exercises as they are designed, not only for practice but also as an additional source of information that will end the hands. You'll be I problem solving skills. Are you excited? Awesome. Let's begin this journey together. See you they 2. Why are there so many Jagons in Business and Data Science?: So what is it about data that is so important in this day and age? Maintaining a successful business goes hunting hard alongside walking with data. Whether you understand it or not, there is no denying that data is the foundation off any successful business on the business entrepreneurs that are living the way away. Looking deeper into data is what will make their most successful. Approve the competition. Let's start with the data team or the big data team. They will want to solve a business problem. That same would do a significant amount off on the data that's available. First business in that's the business intelligence team would provide a business insights. Dashboard rose. The dashboard is ready. They did. The science team will use some business analytics or Data Analytics tools to develop models that could predict future outcomes. Lizzie did the science genius. You may think we're just picking words out of the dictionary and sticking them together at a rondo. Nope. These are actual data science jungles. And not only that, there are many other similar phrases. No wonder you're confused. But don't be. It is completely understandable for you to feel like this little share some light as to how things seem so complicated no one calls off. This confusion is a constant evolution off the data science industry and in turn, the 1,000,000 off these jugglers. This complicates the situation a lot. For example, someone who had it I to lost statistician 25 years ago would have been responsible for Ghadry on Clinton data sets. Applying various statistical metals to the data after some years are we work with the group's off data on a radical improvement of technology, This statistician would now be required to extract patterns from data and sports. In New Jagan was coined data mining. Similarly, a few more years, the same statistician due to new mathematical statistical Bodo could now perform more accurate forecasts on again. Another term US find its way into on already inflated business glossary. Predictive Analytics As the statistician changed a job by this point. Nope. I goes different. Nope, not really. However, she is more qualified now to be part of the statistics department Predictive Analytics team , or are the title data scientists. Hopefully, it is easy to see how this juggles develop over time and how someone would qualify as a statistician 25 years ago on that kept up with modern technologies could feed into the multitude off professional categories. Now interesting but a little messy, right? Another course of confusion, which stands for the one I just mentioned, Elia from it are managers who are understandably, can become overwhelmed with a barrage of new terms and juggles flying around. This causes them to liberal job positions inaccurately. One hr representative may call it your position data analytic specialists, when in fact they need a data analyst. Another may employ a junior data scientist when he require it business intelligence analyst . Of course, there are many companies that want their job offers brilliantly, but this is no standard across the board, which can cause even more of a mess. Now ask exemplified already, the world of data and data science can seem overwhelming. I mean, very well make you want to run away and hide from anything. Dates are related. So in this lesson will begin by clarifying the similarities and differences between the terms of business analytics, data analytics, data science, business intelligence on machine learning, there were focused on helping you digest the definitions you need to know in an effective way. By hand of this Siri's you will be modern, capable of being able to relate on applied various expressions on juggles to the areas of data science they belong to. So what are you waiting for? Less dive straight in. 3. Analysis Vs Analytics: all right, so let's discuss the not so obvious differences between the Terms analysis and analitico X due to the similarities of the world's. Some people believe the share the same meaning, and there's use them interchangeably. Technically, this isn't courts. There is, in fact, a Dixons difference between the two, and the reason for one off from being used instead of the holder is the lack of the transparent understanding off booth. So let's clear his hope, shall we? Frost will start with analysis. Consider the following. You have a huge data set containing data or various types. Instead of tackling Ian, targeted sets are running the risk off the common over warmed. You separate it into easier to digest chunks and started them individually and examine how the relates to the older parts. And that's analysis in a nutshell. The only important thing to remember, however, is that you perform analysis on things that have already happened in the past. Such us using on analysis to explain why a story handed the way it did or other was a decrease in sales last summer. August means that we do analysis to explain how and or why something happened greets. Now this lead falls nice later. The definition of politics, As you have probably guessed. Analytics generally refers to the future instead, off a splendid past events, it explores potential future ones. Analytics is essentially the application of logical and competition, our reasoning to the components parts obtained in an analysis. And in doing this, you're looking for patterns in exploring what you can do with them in the future. Yeah, analytics branches off into two areas. Qualitative analytics. This is using your intuition and experience in conjunction with the analysis to plan your new business move. Quantitative Analytics. This is applying formulas and algorithms to numbers you have got out from your analysis. Yeah, a couple of examples say you are owner off an online clothing store. You are ahead of the competition and have a great understanding of what your customers needs and one's heart. You've performed a very detailed analysis for women's clothing articles are flu sure about which fashion trains to follow? And they used this intuition to decide which style of clothing to start selling. This will be a qualitative analytics, but you might not know when to introduce the new collection in that case, relying on past sales data and user experience deter, you could predict in which months it would be best to do that. This is an example off using quantitative analytics. Excellent to backtrack a little, you can combine these areas with analysis. Also, you could perform qualitative analysis to explain how or why this story ended the way it did. And you can perform called status analysis. Walking with past data to explain how sales decrease last summer. Perfect Now that we have cleared up the differences between analysis on analytics, it shouldn't be too difficult to see how terms such as data analyses, detail politics, business analysis, Business analytics can hardly a unique meanings so more this will be explained in the next lesson, which aims to simplify this as well as many more with a diagram. So let's carry on 4. An Overview of Data Science Task: So how do you approach in business or editor? Science task? Yeah. Two possible scenarios. Imagine your books as carefully ready confidence reports and dashboards. I want you to make some predictions for the firm's outgoing costs over the next year. So the logical way to approach this problem is to get us on relevance data on, Then prepare it for analysis. Alternatively, your boss might come to you and say, Hey, we have this in normal states are here. We don't really know what we can do with it. Its customer data. So it must be useful. Can you do something with it and tell us how we could increase our profits for next year? In this case, I'm in the data set is the starting point for, like, the force case, you don't need to call it data that will help you answer your business question. You provide already data on off you go. You can analyze its on, apply different political tools to extract insight and make forecasts. In either scenario, from a data scientist perspective, the solutions, every task begins with having it proper data sets. This must be forced her to do list. Only then can you proceed with further analyses and forecasting 5. Traditional Methods of Applying Data: okay as a singles thing big, but starts more right. So let's that's more while we think big and requested business completing data data Eyes define us information stored in a digital format, which can then be used as a base for performing analysis and decision making. There are two types of data. Traditional data on big data. Dealing with data is a four step when solving business problems or researching, so it's important to know what you're looking at when you hear the term data. The image that probably pops in your head is what is considered traditional data data. In form of tables. Continue numeric or text values data that is structured on data stored in databases, which can be managed from one computer. Data on the order hand is a whole. Another story on its could be assumed by the name big guitar. The term, reserved for extremely large data on it is not just a mongers. In terms of volume, these data could be in various for months. Its calm, restructured, semi structured or structured big hitter is just That's pink. You also awful. See it characterized by the little V as in three off big data. Sometimes we are five, several or even a level V's off big data. They may include division you have for big data value, big data, Curries, de visualization tools. You use all the variability on the consistency off big hitter. And so, however, the following are probably the most important criteria. You must remember volume. As we already said, Big Deter needs a whopping amounts off memory space. Typically distributed between many computers, its size is measured in terabytes. Better bites on evil exabyte variety year. We're not talking just about numbers and text. Big deter Awful implies the load images. Audio falls. Mobile dates are on orders velocity while walking with big Deter, whose goal is to make destructive pattern from eat as quickly as possible. The progress that has been done in this area is remarkable outputs from huge data sets coming retrieved in real time. This means they can be extracted so quickly that a result could be obtain immediately after the source data as being obtained, which ever type of data you have on your hands it is your first port of call for business problem solving, so it's important to know what you're dealing with it. Now let's move on. So when you have Goddard unorganized all the data, it's time to get your hands dirty With data science on analytics. Data science is a broad subject. It is our interdisciplinary field, the combined statistical mathematical programming problem solving on data management tools . So as any good data, scientists will see less segments. The process Further data science can be divided into three segments. Business intelligence, traditional methods on machine learning. The frosting off. Applying data science is to un alive the past data that you have acquired Business intelligence is a discipline you need for this. We are includes technology driven Jews involved in the process of analyzing understanding, reporting available past data. As a result, in your having reports or dashboards, I will help you on your way to making an informed strategic and tactical business decisions . This part of the process is water. Your time. You can extract insight on high ideas about your business that will help it grow and give you on age off your competitors, giving you a dead stability business Intelligence means understanding how your sales grew on why the competitors lose market share. Was there an increase in the price of your products? Or did you sell in? Mix off more expensive products? What did your profitability margin behave in the same time frame of the previous year? Was their clients accounts that way more profitable? This is what BR is all about. Understanding past business before most in order to improve future performers. Charles, like a great starts right now. It's time to shoot our way. NUS. Up until now, we've been stuck in the past vocalizing previously cleaned information and using business intelligence to explain it past events. But now it's time to look to the future. I'm going to be able to forecast our future sales and profitability. Also will US expenses was your B I reports and dashboards are completed and presented. It's timeto apply one of the two types off data science traditional methods, also called traditional data science or machine learning to develop an idea of what will happen in the future. Traditional methods are calling toe are from work. I set off metals that are derived mainly from statistics on our adopted for business. There is no denying that this conventional data science tools are absolutely applicable today. They are perfect for forecasting future performance which create accuracy to do really idea . Yeah, some instances question analysis. Closer analysis on fractal analysis, all of which are prime examples of traditional metals. Cool. Now let's focus our attention to machine learning with machine Lenin. In contrast to traditional methods, the responsibility is left for the machine through mathematics is significant amounts of complete a bar on applying AI. The machine is given the ability to predict outcomes from data which have been explicitly programmed to machine. Learning is all about creating algorithms. The let's machine received data before calculations. UN. Apply statistical analysis in order to make predictions with unprecedented accuracy. In our let's listen, we will explain where we focused on these five discipline. Precisely. So stay tuned and thanks for watching. 6. Traditional Data Techniques: so in this lesson will reveal what the clicks are used when walking with traditional data. Big data be I traditional detail Science. NATO's on machine learning wells with cover that's who provide examples off where this necklace can be applied in your life and trendy food of data science mean Ulan currents on the terms that can be used. Such asked data cleansing reports, dashboards, metrics, came ins de planning and so on often is essential that you are able to relate each off them appropriately. However, don't worry, as this lesson aims to do. Just that's provide you with definitions. Visuals are realized examples in this lesson who focus on approaches and techniques specific to walk in with traditional data date size. Very broad seem to confer to row fox processed data or information to make sure we're on the same beach. And to better understand the different techniques that apply to a deficit will separate those phrases. Take a look at the following graph People gather raw data that must be processed to obtain 1,000,000 full information. Great raw data, also called raw funds or primary data, is detail which cannot be analyzed the way it is going touch data you've accumulated and stored on the server. The garden of raw data is referred to ask data collection. This is the first step data can be collected in several ways. One example would be the use of Soviets asking people to read our modesty like or dislike a product or experience on a scale of 1 to 10. Alternatively, Guardian data could be automatic cookies, a common example. Off this, they provide companies with detailed information about a user's activity on the website. Put examples can give companies a good idea of how to help improve customer satisfaction, but not yet. This data can have various programs we need before it's off any use. Therefore, it's most be persist. Okay, good. Now let's look at how the process in step take place and can turn raw data into useful and meaningful information. The first thing to do after a godly raw data is what is called data pre processing. This is a group of operation that will basically convert your road data into a former that it's more understanding on, hence useful for further processing. So let's put this tool between raw data processing on our graph So what was it a pre processed aimed to do exactly its attempts to fix the problem That can, inevitably. A call with data Ghadry a problem. For example, with some customer data that has been carded, you may have the person registered asked. 932 years old or say United Kingdom, as the name of a person obviously does. Deter entries are incorrect. Before proceeding with any type of analysis, Salzgitter must be marked as invalid or corrected. That's what data pre processing is all about. Perfect. So let's dive into specific techniques that are applied while prepossessing raw data. One technique is class labeling. This involves liberal in the data points to the correct. It's a type or arranging detail by category, while such category is in a miracle. For example, if you're sternly number off goods sold daily, you're keeping track off numerical values. These are numbers which can be money politic, such as the average number off ghouls sold body or both. The older liberal is categorical. You're dealing with information that's cannot have a mathematical manipulations, for example, a person's profession or place off beds as this just provide information about them. I don't think making need to be familiar with these dates are cleansing. Also, loner states cleaning or deter scrubbing The goal of data cleansing is to deal with inconsistent data. This girl come in various for, so you are provided with a data sets containing the United States. On 1/4. All the names are misspelled in this situation. Sentence techniques must be performed to correct these mistakes missing that are another thing you'll have to deal with. That's all customers would give you. The data you are asking for what often happens is that the customer will give you his name on occupation, but not the age, for example. Now what do you do? In that case? Should customers entire called be disregarded? Or perhaps you should just enter the average age of the remaining customers. It's a confusing on dealer with missing values are problems that must be solved before you can prove, says the data for the great. Now let's move on to a couple off approaches that are specific to walking with. They should not data the 1st 1 being balancing. Imagine you have compartments Lovie together data on the shopping habits, off men and a woman he went to Los Attaining will spend more money during the weekend. For example, however, when you have your data, you will notice that 80% of respondents when women on only 20% meal the training amenities are not going to be two worlds. Men as much healthy are two women to cancer at the problem. Applying balance and technique will be the best thing to do, such as taking equal number off respondents from each group, so the ratio is 50 50. Another common approach is data shuffling. Shuffling. The observation from the data set is just like shuffling a deck of cards. It will ensure your data is free from unwanted butters caused by problematic data collection data. Shuffling is a technique which improves predictive performance on else. Avoid misleading results. But I don't think avoid elusive results. Well, it is a detailed process, But in a nutshell, shuffling is a way to randomize data. If I take the 1st 100 observations from the deficits, that's not a random sample observation that we entered first, who were forced to be extracted. If I show for the data, I'm sure that when I take 100 consecutive entries there will be random on most likely representative. Finally, I would like to show you two visualizations frequently associated with databases containing traditional data or, in a more technical language, relational database management systems. The First World represents the entity relationship diagram or E. R. Dag. This accomplice theoretical way off illustrates in a database architecture specialist. Love it for the streets, Lord we in which the different ships going out tables in the database are related. The second Google is important to be understood well, not only by database creators but also by X users. It is called relational schema. Here, each rectangle represents a dick stint detectable on the line. Shows which tables are related are which aren't to record data collection. Unprepossessing are essential for quantitative analysis. Now there are many more techniques that can be applied to trash. In our data, we wanted to show you the most frequently used techniques out there. Thanks for watching 7. Traditional Data Real Life Examples: Okay, so let's provide a couple of traditional data examples. Thank old bitty customer data. This table contains text information about a Guinea customer. They're going to use it to give a clear example, or the difference between in a miracle on categorical variable, not just the force column. Shoes I designed to the different customers taste known was, however, cannot be manipulated. Calculating an average idee is not something that will give you any sort of useful information. This means that even though their numbers, they hold no numerical value and therefore represents categorical data now focused on the last column. This shows how many times customers filed a complaint. These numbers are easily manipulated. Added ample together. To give it to the number of complaints is useful information therefore under medical data. Another example. We can nuke heart is daily historical stock price data. Any data set? You see, there is a column concerning the dates off the observations, which is considered categorical data on a column containing the stock prices, which is numerical data great. Additionally, we consider these two examples were representative off traditional data, structurally doing also complaints or too large to be organized into tables on detail. We Claire on understandable 8. Big Data Techniques: some of the approaches used. Untraditional data can also be implemented on big data. Collecting a process indicator is essential. So help organize the data before doing any analyses or making predictions as his grouping data into classes or categories. While walking would be detail, things can get a little more complex. You have much more variety beyond the simple extension off Miracle on categorical data. Example. Big deter can be text data Digital image data. Did you tell video data? Did you tell or do data on more? Consequently, with a larger amounts of data types comes a wider range of data cleans and Melton's. There are techniques that verified that a digital image observation is ready for processing . A specific approaches exists. Ensure the audio quality off your fall is adequate to proceed. So what about dealing with missing values? This step is a crucial one, as big data has big missing violence, which is a big problem to exemplify. Let's take a look at some clear specific techniques for dealing with big data texted her mining represented process off deriving valuable on structure detail. For me text, let's elaborate. Think of a huge amount of text that is stored in a digital formats. Well, there are many scientific projects in progress, which came to extract specific text information from digital sources. For instance, in the RV database, which are stored information from academic papers about marketing expenditure, the mental pick off your research. You could find the information they need without much of a problem you know will seize on. The volume of text stored in your database was low enough. Often, though the details huge, it's may contain information from academic papers below articles, online platforms, Private Excel files on more. This means you will need to extract marketing expenditure information from many sources in all the world's big data, not on it. Attacks, which has led to academics and protection. ALS developing matters to perform for US states are mining who? What else? Data Maskin. If you ultimate in incredible business or governmental activity, you must preserve confidential information. However, when personal information is shared online, it doesn't mean that it can't be touched or used for analysis. Instead, he must apply some data masking techniques so you can analyze the information without compromising bribery. It's details like glitter show Flynn that's a mosque and can be quite complex. It conceals the original data, which random on false data, allowing you to conduct analysis and keep all confidential information in a secure place. An example of applying data masking to big data is true. What we called confidentiality preserving data mining techniques. Nice. Listen, look at this graph. It's a simplified version off what big hitter teams, because simply dealer with in contrast to tradition, are visualizations. The object in there represents data tables under inside relationships, rather to shoot a complex way in which data derived for many sources. Physically, whenever you see a large number of lines intersect in, most probably you're looking at a visualization about big data the hotel to examine the graft orally. You can appreciate how much other it is to make sense or big hitter compared to treasure now, data. Okay, 9. Big Data Real Life Examples: all right, So where do we find big? Deter The size and increasingly more industries on companies? Here are a few notable examples As one of the largest on my communities, Facebook keep tracks off its users names, personal data, photos, videos, recorded messages. And so this means that their data has a lot of writing in it. With over two billion users Ward Wild, the volume of data stored on their servers is tremendous. Facebook requires real time reporting off the aggregated and anonymized voice off its users on it applies many analytical tools for its mobile applications. This munity confident is investing in posting its real time data processing powers. In present, velocity off its data sets great stuff, Know what else can we find? Big data, domestic financial trading data, for example. What happens? We will record the stock price every five seconds. For every single second we get a data set that is incredibly volume in years, requiring significantly more memory dicks space in various techniques to extract meaningful information from it. Little like these would also because he had big data. We hope this example terrify the use of big hitter in the real world. Thanks for watching 10. Business Intelligence Techniques: So let's assume your data has been pre processed on Is ready for analysis. It is not wrong anymore. It is beautifully organized, and as you go through it, you can understand the information each entry conveys. Great well. D analyst. This means you are ready to enter the room off business intelligence. This is the point where all your deter skills combined with your business knowledge and intuition to explain the past performance off your company with conviction. It would confidently be ableto as our questions, such as what happened When did it happen? How many unity to the cell in which region did we sell the most goods on many more. A good business intelligence analyst does not look at a problem. For one single angle, she will be able to extract any and all information required to solve. This problem should be capable of answering simple questions from our manager, like Western News regarding our in the Rocket and as well as more complex ones like how did our human market and perform last quarter in Terms off click through its revenue generated ? And how does that compare to the performance in the same quarter last year? the drop of a business intelligence analyst is unimportant. Want an interesting one? It's a job that requires her to understand the essence off the business and straightening that business through the power off data. Amazing. Let's look at B I from a technical viewpoint on what we are. Seat. Let's introduce some relevance PR terms. Okay, so how do we measure business performance? We start by collecting observations. For instance, you can observe variables. Such a sales value or new customers were enrolled in a website. Each monthly revenue or each cost tomorrow is considered a single observation basis, all grades. However, no mathematical manipulations can be applied to these observations. What we must do is quantify that information quantification is a process off representing observations asked numbers. For instance, if you see your revenues from new customers for January, February and March, where on dread? 1 $2130 respectively, while the corresponding number off new customers for the same three months are 10 15 and 25 fantastic. Next, let's discuss nausea and measure is the accumulation of observations to show some information. For example, if you had to total revenue off all three months to obtain the value of 3 $50. That will be a measure off the revenue of the first quarter off that year. Similarly, act together the number of new customers for the same period. 50 and you have another measure. Great. They order frequently used terms. Do you know what in metric is in metric refers to a value that derives from the mergers you obtain on aims at gauging business performance or progress to compare if the measure is related to something like simple, descriptive statistics off past performance in metric as a business many attached. Let's exemplify this. If you estimate the average quarterly revenue, the new customer, which equals 3 50 divided by 50 that is $7. This is in metric. Once again, it is not just a number, but in our mother. Beers in business a minute metrics off the readers who for comparisons. For instance, you contract the average quarterly revenue per new customer every three months on those computer customer retention. On spending every quarter in it, not show the observations. Least two measures are your measures I used to create your metrics. Your metrics are business applications. Nice. This is all fine However, in the real business, with the number off observations is significantly larger. You can derive hundreds or sometimes even thousands off metrics. Can we keep track off all possible metrics we can extract from a data sets? Probably yes doesn't make sense to do that. No, what you need to do is choose in metrics that are tightly are lined with your business objectives. This metrics are called KP eyes Key Performance indicators key because they are related to your main business schools performance because they show are successful. You have performed within a specified time frame on indicators because their values or metrics indicate something related to your business. Performance, for instance, in metrical indicates the traffic off a page from the website that was visited by any type of user, a k p. R. Instead, we showed the volume of the same traffic, however, generated only from users who have clicked on the link provided in your heart campaign. Those you could check if the heart your sense I source of motivation for customers to click on it over the link and get the specified page of your website, which will determine whether you continue to spend on the hards or not. Also examine measures are calculating metrics on KP eyes are not. So rule off the business intelligence analyst. Every con stated, meaning you can extract most be visualized general managers will want dash boats are reports with graphs, diagrams, mops on other easily digestible visuals supported by most relevant numbers. Fruit tree Now the boring metrics and turning the interesting and informative KP eyes into a easily understood and comparable visualizations is an important part off the business intelligence analyst job. Often, dashboards are created about KP eyes only Keep your dashboards looks similar to the world's containing metrics. However, they allow executives and managers to stay focused on the main business objectives. Of course, you will able to create such dashboards later on in the course series. This was all we have to see about business intelligence. What I'm being Thanks for watching 11. Business Intelligence Real Life Examples: So let's see a couple of examples that prove that business intelligence is not only for in theory, but also in practice. Did you know that be I can be used for price optimization Hotel Use price optimization very effectively by raising the price of your room in periods when many people want to visit the hotel and by reducing it to attract visitors. When demand is lou, they can greatly increase your profits in order to competently apply such a strategy. It most extract the relevant information in real time and compared with historical being, allows you to adjust your strategy to past data. A solar sits available. Awesome. Another application of business intelligence is an ass and invent Torri management over on undersupply course problems in the business. However, implementing effective in Venturi management's means supplying enough stock to meet demand with a minimal amount of waste and costs. To do this well, you can perform on Indian analysis off past sales transactions for the purpose off identifying seasonality buttons and in times of the year with the highest sales. Additionally, you could track your inventory to identify the months off which you have over or under start it. Detailed analysis can even pinpoint the day or time of the day when the need for your good is highest, you don't write business. Intelligence will help to efficiently manage your shipment logistics and in turn, reduce costs and increase profits. 12. 7 Traditional Methods Techniques: All right. So you've prepared be I reports and dashboards on. The executives have extracted insights about the business. What do you do with information? Use it to predicts all future values as accurately as possible. That's why at this stage you stop dealing with analysis. I start applying analytics not precisely predictive analytics. Remember that was separate predictive analytics into two branches. Traditional metals, which comprises classical statistical methods for forecasting on machine learning in this lesson will focus on the first type. So I started technics involved when applying traditional deter science metals. So there are many. It would be often come across is regression. But what women by regression and business statistics. The question is, in more they'll use for quantifying causal relationships among the different variables, including your analysis. Let's suspend this drawn example. First, look at this data sets in one column. We have house prices in dollars, while any other outside is measured in square feet. Every rule only data table is an observation on each can be plotted on this graph as adults . Outsize isn't measured along the results online on its price on the vertical line, the photo to the right on observation is a larger, the outsize on the father hope the higher the price. So once we plotted all 20 observation from our deficits are graph will appear like this the Phillies There was a street like old in Grecian life that goes through these dots helping us close that it's gonna be toe all of then simultaneously. Now imagine would re on our line like this. Do you think the observation altogether? A closer to the foster in line on the second green one? Yes, they're closer to the red line. This means it more accurately represents the distribution of the observations. So we don't getting too technical in this lesson in my line, this is associated with a specific mathematical or asked Irritations will see econometric expression in this case if why signifies the house price and B represents a coefficient which will multiply by X outsize. So the question professional walk quits is what equals b Times X and they used a graph US visual support. Now let's concentrate on the visualization and leave them automatics for leader. Excellent. So, naturally, if there are linear regression models, that must be nonlinear worlds as well. True corrects a logistic regression is a common example off in northern India. Rodeo. In this case, a lot of the house prices examples values on the vertical line won't be obituary inter. Just there will be words for zeros only. Such a model is useful during in decision making process. Let's elaborate on that. Corbyn's apply logistic regression to future job candidates during the screening process. If the algorithm estimates the probability that a prospective candidates will perform well on the conference are both 50% it would predict one or successful application otherwise to predict zero. Therefore, the nonlinear in nature off logistic regression is nicely summarize by its graph very different from the linear regression rights. Okay, grits, are there different types of aggression? Look at this graph on again later to the house prices. Example. This observation, called from 1/3 date, is it? Imagine the are derived from research on German house prices. Hence, there dispersed differently should control the regression line. What? Could we do something better? Yes, we could. When he did, I divided into a few groups called clusters. You can apply closer analysis. This is another technique that will take into account that setting observations as it beats similar house sizes on prices. For instance, this cluster off observations denotes small houses are pretty high price. This could be typical for houses in the citizens on the second post. Op will represent houses that are far from the city because they're quite big but cost less . Finally, the last most Arkansans houses that are probably not in the city centre but are still in nice. And the roads they are big cost a lot. All right. Notice Entire data can be close. That is important so you can improve your for the analysis. In the example close turn allowed us to conclude that location is a significant factor when pricing a house. Great. And what about a more complicated story where you consider explanatory variables? Apart from how size you might have quantified the location, number of rooms you have construction and so on. They can all affect house price. Then one thinking about the mathematical expression corresponding to the regression model. He wouldn't just have all explanatory variable x. You will have many x one x two x tree and so on. No got an explanatory variable can also be called Eric Wrestle and independent variable or it predicts or variable. Now let's move here for our surprising example. Imagine analyzing SRV that consists off under questions, performing any analysis on 100 different variables stuff. This means you'll have very well starting from X one on going all the way up to X 100. The goofiness that's often different. Questions are measuring the same issue, and this is where factor analysis calm scene. A senior Soviet contained this question on a scale from 1 to 5. How much do you agree with the following statements? One. I like animals, so I care about animals. Three. I'm against animal cruelty. People are likely to respond consistently to these three questions. That is, who have a marks. Five for the first question does the same for the second and the third question as well. In other words, if he strongly agree with one of these three statements, it wouldn't disagree with the order to rights. Well, with factor analysis, we can combine all the questions into general attitude towards animals, So instead of these three variables, we can have one in a similar manner. You can reduce the dimensionality off the problem from 100 variables to attain, which can be used for regression that would deliver in more accurate prediction. Fantastic. To sum up, we can see clustering is about grouping observations together on factor Analysis is about grouping s planetary variables together also, Finally, I would like to show you what I'm serious. You use this technique, especially if you're walking in economics or finance in these fields, you would have to follow the developments offset in values over time, such as stock price or sales volume. You can associate time Siri's with plotting values against time. Time will always be on the or results online as time is independence off any other variable . Therefore, such a graph can end up depict in a few lines that illustrates the behavior off his talk over time. So when you start individualization, you can spot which talk performed well on which did not excellence. We must admit there is a variety of metals that professionals can choose from. This was just a brief introduction into the basic types. Keep the pace for the next lesson. Thanks for watching 13. Traditional Real Life Examples: So when will you deal with traditional statistical methods? Application of correspondent Technics is extremely broad, however, will provide you with two examples that will help you associate some techniques that were presented any previous lesson to reward instances. Imagine you are the head of the user experience department off the website, selling goods on a global scale which we often aggravates us. Your ex. All right, so what's your goal as head of UX? To maximize is a satisfaction, right? As you may have already designed, unimplemented is so V that measures the attitude of your customers towards the letters global produce. You have launched the graph for you plotting observations were likely up here in the following way. No details concentrated in such a way Issued cost are the observations, remember? So when you perfect cluster analysis is, you find out which costs are represents different continents. This group may refer to the responses guarded from Asia, this one from Europe, this road from South America on the last one over a year from North America. Once we realize there are four Texans groups, it makes sense to run four separate tests. Obviously, the difference between Klosters is too great for us to make a general conclusion. Show me. Enjoy using your website in one week. What Europeans in another those It would be sensible to adjust your strategy for each of these groups individually. Greats. Another noteworthy example we can give you is forecasting sales volume. Every business and financial complaint does this. So which traditional statistical technique that we discuss? What 50 picture here? Time series analysis. It is said, This is your data on Syria. Set indeed, What will happen next? How should you expect to sales to be for the next year ahead? Really? A volume increase or decrease? Several types of mathematical and statistical models allow you to run multiple simulations , which would provide you with future scenarios. There's on this scenarios. You can make that our predictions on implements adequate strategies. Also, you already acquainted with many of the deter science essential attempts, but not whole. Not yet, and the next lesson will present the most important thing you need to know about machine learning. Thanks for watching 14. Machine Learning Techniques: Okay, let's talk about the other branch off predictive analytics on the last one to be discussed . Machine leading. So the question is, in what situation is it preferable to make predictions? Using machine learning rather than traditional statistical methods in this lesson will provide you with an example showing you what machine Lenin is all about. Grit less begin what is at the core off machine learning created an algorithm, which computer then uses to find a model that fits to deter as best as possible makes very accurate predictions based on that. And how is that different from conventional metals? Well, we don't give the machine instructions on how to find that Modo provided with algorithms would give the machine directions on how to land on its own. So how can we describe emotional in our garden? In a few words, In machine learning algorithm is like a tryout on arrow process. But a specialty in about it is that each consecutive TROIA is at least as good as the previous one. Technically speaking, there are four ingredients. Data Modell. Objective functions on optimization are going in. Let's illustrate them with a fictional example. Imagine a robot old in a bowl. We're going to find the best way to use that ball to fire accurately. In other words, the usage of the bull is Armado. The best way to let after is to train right. We trained by taking different arrows on trying to eat the targets. So takeover of the Haro will be our data or, more precisely, the detail that the robot who used for training. They are all arrows. Will they have. They are supple. It is. There are street ones who could wants light ones on everyone's. So we can safely say the high rows represent different data values. Okay, we said the robot will be firing at its targets, right? Well in machine Lenin, or at least in the most common type, supervised learning. We know what your human for our we call it. That's right. Eight targets. Here comes the third ingredients. The objective function objective function will calculate out far from the target the robot short with on average, here comes the fort ingredients. The optimization are good in it steps on finding off. The objective function consists off the MEC Knicks that will improve the robots actually skills. Somehow it's posture the way it's Moldable, how strong it's pulls the ball, string, etcetera. Then the rebels will take the exact same data or arrows on fire them once again, which its adjusted posture this time they should still be on average, grows out of the center on the target. Normally, the improvement will be almost unnoticeable. This entire process could have been hundreds or thousands of times onto the robots finds the optimal way to find the set off hours on each the center every single time. The machine learning concept in this example is clear. Nevertheless, it is important that while training he would provide a robot with a set of rules that issue will have programmed. It sets off instructions like Play zero in the middle of the bull portable string and so on . Instead, he would have given the machine a final go place de Haro, in a sense of how the target, as you don't care if it's places de Haro in the middle or in the bottom of the bowl, so long as it's eat target. Also, I don't know. Important thing is that it will lend out shoot well right away. But after a 100,000 tries, it may have learned how to be the best archer. Are they, anyway? Now you might be asking yourself, are in Their infinite possibility is to try a one with the robots. Stop training most interestingly, the robot to lend seven things on the way. I will take them into consideration for the next shot at fires, for instance, even close that it's most looked towards the target. It will stop firing in the opposite direction. That's the propose off the optimization. I'll go with them. It's gunfire horrors forever. However, it in a sense on nine out of 10 times maybe good enough so we can choose to stop it after it reaches is set a level of accuracy or fires. It's setting number off our hours. Nice. So let's follow. The four ingredients at the end of the training are robots or model is already trained on this data. With this set off pharaohs, most shots eat the center so they haro or the objective function is quite low or minimized . As we like to say the posture, the technique on all other factors cannot be improved. So the optimization are Gordon as don't it's best to improve the shooting ability off the machine. Great, the only robots that is an amazing archer. So what can you do? Give it a different bag of arrows if they had seen most type of horrors while training, it'll do great with the new ones, however, Forgive it after narrow or a longer hard on, it had seen its will know what to do with it and all ordinary cases due respects the ruble toe if dissenter. What least get very clues? The benefits of machine learning is not robots. Carlin to fire more effectively done a human it might even discover will be holding bulls in the wrong way for centuries. To conclude, you must say that machine learning is not about robots. What people use it for is to improve complex competition on models that conferring infinities applications in our daily lives, especially in the business ward and that's complex competition are models were talking about steps on the fundament. It's off regression cluster analysis models who discussed in our previous listen and take them to unconventional but extremely useful divorce. In our next lesson, we'll discuss that three types off machine learning and provide two examples of reward applications off machine learning. See you there 15. Types of Machine Learning: There are three types off machine Lenin, the most common one applied in a majority of cases we call supervised Lenin, which the robot archery example was referring to in the previous listen. Its name derives from the fact start training on algorithm resembles it teacher supervising her students to a liberates apart from the front ago is set to the robots. It's important to mention you have been dealing with a liberal data. In other words, you can assess the accuracy off each shots. In fact, there isn't a single target. Different arrows have their own targets. Clarify. Let's check what the robot sees when shooting the ground. It's I got at a short distance. It's I got ready for the distance. It's I got hanging on the tree far behind. It's a house to decide on a sky. So having labeled data actually means the following. Associating or liberal in its are get to a type of horror. You know, that was a small Haro. What is supposed to it's the closest star Get with the medium Haro. It's gonna rigid targets. Get it for thy away. Well, with a larger Haro, it's I get that's hanging on a tree. Finally it cook it. Haro is supposed to hit the ground, not reaching hen inside it. During the training process, the results will be shooting Harrell's not respected targets as well as it come after the training is finished. Ideally, they were about to be able to fire the small Haro at the center of the closest target, the Middle Haro, out of center off the one for thy away and so on to summarize. They know that that means we know that I get prior to the shots on. We can associate that short with its are get this. We were sure where the Haro should hit. This allows us to measure the inaccuracy of the shots through the objective function and improve the way the robot shoots through the optimization algorithm, something we explained in the previous lesson in more detail. So what we supervise is the training itself. If it shot is far away from its target, we correct the post your otherwise we don't great in practice. Do. It might happen that you won't have the time or resources to associate the horrors with targets before giving them to the robots. In that case, you could apply the other major type of machine learning all supervised learning. Here. He would just give you a robot. A bag of pireaus, which are known physical properties are labeled data. This means need a you know, the robot will have separate at the high rose into groups. Then you're actually machine to simplify our in a direction which are providing it which targets. Therefore, in this case, you want to be looking for a model that helps you shoot better, brother and we're looking forward, which divides the Haro in a certain way. Here is a quick overview off what happens? The robot will see Chelsea ground the tree, the house on the sky. Remember, there are no targets. So, after firing thousands of shots during the training process, will end up having different types of arrows stock in different. Here's for instance, you may identify all the broken horrors by noticing they are falling on the ground nearby. The orders, you may realize, are divided into small, medium large Barrows that maybe anomalies like crossbow bolts in your bug that after being shot, they have accumulated in a power over you you wouldn't want to use them with this simple ball, would you? At the end of the training, the robot will have fired so many times that it could discover answers coming. Surprise you. The machine may have managed to speed the higher ALS not into four into five size categories due to discover in the crossbow bolt. Or it may have identified that some heroes are going to break soon by placing them in the Brooklyn Haro pile. It is what's mission in that's the provides land and can deal with such problems to any toast. That's very often, However, if you have one million ha rose, you don't really have the time to assign targets toe all of them, do you? To save time and resources, you should apply unsupervised learning. Another important thing so hard is that in practice are supervised as provides. Lenin may be applied on his hand, taking advantage of the different questions they're supposed toe Hansa. In our example, you might want to leave your bag with many horrors of all clear, ah, roadsides to the robots and just let it shoot until it speeds. Your dates are into five categories or more precisely five clusters, then 90 types of data you have available. Even fire a few shops with the different types yourself. Those figuring how to targets each type of Haro can heats. Finally, you can give the already close that data set to the robots, providing it. Which targets are letting it shoots onto your pertain in more precise model for using the bow that would be supervised. Lenin Great, they thought. The media type of machine learning is called reinforcement learning. This time we introduced a reward system every time the robot fires an arrow better than before it to receive on our word, say chocolates it to receive nothing if it fires worse. So instead of minimizing on a row, we are maximising in reward or, in other words, modernizing the objective function. If you put yourself in the shoes of the machine, you will be reasoning in the following way out fire hand Haro and receive your reward. I'll try to figure out what I did correctly, so I get more chocolates with the next short or I fire an arrow on. I don't receive a reward. The most is something I need to improve for me to get more chocolates on my next shots, Steve reinforcements. Also, in addition, don't forget the robot Archer was on abstract depiction of what immersion model can do in reality. There are reports, yes, but the model will be a highly complex mathematical formula. The heart will be a data set on the goal, with materials on quantifiable here, the most notable approaches you'll encounter when talking about machine learning. Support vector machines. Neural networks deplaning random forest models on Bayshore networks are all times off supervised learning. There are neural networks applied toe on unsupervised type of machine learning, but came in is the most common on supervised approach. We also have deep Lennon, By the way, you may have noticed we have placed deplaning in both categories. This is relatively a new revolutionary competition on our porch, which is acclaimed as a state off the heart machine. Learning today, describing it briefly, you can see it is fundamentally different from other approaches. However, it as it brought particles, cope off application in all machine learning areas because off it's extremely I accuracy off. Its model finally know that deplaning is still divided in supervised unsupervised on reinforcement, so it solves the same problems but any conceptually different way. All right, this was a brief introduction to the vast ocean off techniques that emotional and indiscipline comprises at the moment. In the next lesson will illustrate so real examples of Martian landing. Thanks for watching. 16. Machine Learning Real Life Examples: So how does martial then? And feet into the world of data science? One example is the financial sector, and banks in particular. They have dynamos. It sets off credit card transactions. Unfortunately, banks office in issues which fraud daily they are tasked with preventing fraud. Starts from a crying customer data on in order to keep customers forms save, they use machine learning algorithms. They take past data on because you can't tell the computer what transactions in your history will edge to Mitt. I would were found to be fraudulent. Taken liberal data. Ask such so through supervised learning the train models that detect fraudulent activity. When this models detect even the slightest probability of theft, the flag pretty transactions and prevents the fraud in real time. Although no one in the sector average a preferred solution, the impact off machine learning algorithms has been ground breaking. Another example of fuses by Vice Marshal Lenin with liberal data can be found in clients retention. It focus off any business beats a global supermarket chain on online clothing Shop is to retain its customers, but a largely business grows the other. It is to keep track off customer trends a local shop owner will recognize and get to know dear most loyal customers. They offer them exclusive discounts to take them for the Acosta and by doing so, kick them returning on a larger scale, companies can use machine learning past liberal data ultimate the price. With these, they can know which customers I purchase goods from them. This means the stock on offer discounts on a personal touch any different way, minimizing market costs, a maximizing profits. So I will explain the men taken excuse for walking with data on data Science on the applications will move on to our next topic. See you, they on. Thanks for watching. 17. Programming Language and Software Employed in Data Science: So how are techniques used in data business intelligence for predictive analytics applied in real life? Certainly, with the help of computers, you can basically split the relevant tools into two categories. Programming languages on software known a programming language enables you to devise programs that can execute specific operations. Moreover, you can reuse this programs whenever you need to execute the same action. Our, um, bite on the two most popular Otto's across. Their biggest advantage is that they can manipulate data and are integrated with multiple data and data science software platforms. They are not just you table for mathematical and statistical competitions, in other words, are on bite on our adaptable. They can solve a wide variety of business, and it's all related problems from beginning to the hand. Of course, our fight on to have their limitations. They are not able to address problems specific to some do means. One example is relational database management systems. There SQL is king. It was specifically created for that purpose. SQL, at its most advantages when walking with traditional is to Rico dates. When prepare your be analysis, for instance, you will surely employed okay when it comes to the desires. Mentioning much lab is inevitable. It's ideal for working with mathematical functions or metrics manipulations. That's why it is present in all categories. Except for big data, we're respectable. Much lob usage is a paid service, and that's one of the reasons why it is losing ground to open source languages like our Pitre. Either. We are fighting and much lob combined with ehskyoo, cover most of the tools used when working with traditional data. B I on conventional detail. Science What about big data? Apart from our, um, biting, people working in this area are often proficient in other languages, like Java Scala this to have not been developed specifically for doing statistical analysis . However, it turned out to be very useful when combining data from multiple sources. All right, let's finish off with machine learning. When it comes to machine learning, we often deal with big data. Those we need a lot of competition or power and we can expect people to use language is similar to does in Big Deter column about him are biting on my club. Other faster languages are used like Java JavaScript, C C plus plus on its Carla who what we said may be wonderful, but that's not all. By using one or more programming languages, people creates applications, software or, as they're sometimes called software solutions that are just it for specific business needs . They're small scope doesn't make them less is full. In fact, just the opposite. They're a lot easier to land on, be adopted by orders. You have already heard off several off dues because of its ability to do relatively complex compositions on good visualizations quickly excel. Is it still applicable to more than one category, like traditional data, be I on data science? Similarly, SPS s is a very females to for working with traditional data on applying statistical analysis. Among the many applications we have plotted, we can say there is an increasing amount off software designed for walking with big data such as Apache Hadoop Apache Edge pays on mongo DB. In terms of big data, a dupe is listed as a soft way in the sense that it's a collection of programs. But don't imagine it is a nice looking application. It's actually a software framework, which was designed to address the complexity of big data on its competition. Al intensity most notably a dupe distributed competition, our trust or multiple computers, which is basically the way toe under dictator. Nowadays, Barbie I Sauce on, especially W, are talking of examples off software designed for business intelligence visualizations in terms of predictive analytics. E views is mostly used for walking with Econometric Time, Siri's models on Starter for Academical Statistical and Econometric Research, where techniques like regression cluster of factor analysis are constantly applied as a final note. Remember the following Should you have the relevance business and theoretical knowledge learning software tools relatively easy are supposed to learn in a programming language. More importantly, it to be sufficient for your need to create quick and accurate analysis. However, if you're heretical, preparation is strong enough. You find yourself restricted by soft way. No. Any programming language, such as are on Bytom gives you the freedom to create specific our doctors for each project you're working on. Great. But what would give you a good idea the level of applicability of the most frequently used programming or software tools in the field of digital science? Thanks for watching 18. Data Science Job Positions: it's your managers got also become overwhelmed with the amount of data science terms and Jack goes flying around. This means there sometimes liberal job positions in a misleading way. There's you can end up confused about out of March, a job title with a discipline. So what are the job positions for each activities later? Appetites on Data Engineer on Big Hitter Architects and Big Hitter Engineer, respectively, A crucial titles on the markets it present in these rules is regarded as a very important part of the entire process off solvent data science or in business task. It is an architect. Creates disabilities from Squatch. They designed it. What idiots album retrieved, Processed on Consumed. It's asked off the detail engineer step on the walk off the data architect. His primary job responsibility is to, for the process, the obtained data so that it is ready for analysis. So the results of his work it's something analyst on people in the analytics. A position Will Everly rely on it clean on organized details. It's right in front. The data in it, it's obvious, is not created once and for all you have is setting flow into our from the database, and there is a person who unders this control off data. That position is database administrator, and she really works with traditional data it leads to see administration off. Dictator is usually automated. Fantastic. The FBI analyst would do analysis and reporting off past historical data. What FBI consult on those exactly, is vague. Do be. I consultants are often on external b. I analyst. Many firms also sent it to science departments as they don't need or want to maintain. World B. I consult times would be FBI analyst. I did been employed. However, their job company varied as they hope Holman off difference projects. Finally, FBI developer is a person who undergoes more advanced programming tools, such a spy item on especially sq, in order to create analysis specifically designed for the company. It is the third most frequently incontinent your position in the FBI team over for Nice. Now you must see that the remaining teams are so mixed in the line between the activity off one and the other is very thin. It person will employees traditional statistical metals or unconventional machine learning techniques for making predictions goes to Billy Build a detail scientist Moreover, data analysis is a job title for this report. Pay more advanced types of data analysis under the basic part off the predictions off the data science team Finally, in machine learning Engineer, this job is tough to do, but it's easy to classify. It refers to does who are looking for ways to apply the states off the arts competition. Our models developed in the future of machine learning into solving complex data. Science on business trust are open. Enjoyed this. Listen, thanks for watching. 19. Dispelling Common Misconceptions: hooked up with representation. So far, we have managed to shed some light on money terms on processes related to walk in with data on data science solidified the concepts mation. Let's give a brief summary. I refute some common misconceptions about a few areas on data science job saying 200,000 lines of data constitutes big data. It's simply not true. As we already said, we must consider many things before define it, Deters said. As big on volume is just one of them. Big data is also characterized by the variety off data types involved, the durability of sources. The information has been retrieved from the velocity with which it is processed, and so therefore, don't consider a data set speak just because it feels plenty of rules. When it comes to be. I remember they should be able to distinguish quantitative from qualitative analysis. Not every type of analysis coming considered business intelligence bi ideals with the explanation of past events using data driven approaches on reaching guitar driven conclusions. In contrast, it business analogies approach, which is not data analytics and therefore not be I SWAT analysis, so it's another senses if him off, Stipe off qualitative analysis, contributing to the strategic our decision making of a company. It's points how the strains on the weakness of running a particular business. So it's what analyses can improve the firm's strategy. But it is not a guitar driven analysis. Grips. We can say people often associate traditional statistical metal. It's off coding possible constitutive mathematical are statistical models. However, this is not always the case. Are, um, Pitre. I'm not the only tools you will need if you are working in this sector. Storytelling is a crucial skill for a data scientist. They should be able to have the ability to express complex mathematical and programming concepts to end users such as company manages on high level executives in only a paragraph on a single visualization. To achieve this, software tools such as Excel starter on SPS s are still frequently used. Finally, since machine learning and AI are not particularly old disciplines on because they're tightly woven to the evolution of technology, many developments are still ongoing. As such, these developments are surrounded by disputes from scientists on academics who have not reached firm conclusions. For example, it just well known that deep learning algorithms will let a model perform exceptionally well. However, it is still unclear out the machine can obtain such outstanding results. There are numerous scientific papers looking precisely into that problem, but for the moment, progress on the topic is limited. This was the last stop of our introduction to the world of data and data science. It started off with a specific go to clarify the minion of the most frequently used terms and juggles in the field, and we hope we managed to achieve its. We hope you have enjoyed what in this less is, and I've brought in your overview off the disciplines mentioned Well, that was it would like to invite you to continue with the second series of what top six skills that make a great B I on a list where will actually teach a large part of the techniques and skills may shown in this lesson. We've just begun. Our joining for the time being will never get tired of saying Thanks for watching