 My name is Michael, I'm from here, and it's what I'm making today, it's about making sense to use your quality content to remember myself, what I'm teaching you how to make. So, let me just go straight into what I want to show you about moving, that for all of you, it's something you do not maybe sing or maybe not possible in Microsoft Excel, predictive analysis. So beyond that, there are how many types of analysis in the level of where the analysis comes from, you're going to do a focus on different types of analysis, how many types of analysis, predictive analysis, predictive analysis, any other one. So, predictive analysis, that's in tactics, let me just show you how that, any other one. Don't care about that, doing this might be preventable, but when I explain why the quality certificate is also exploratory. So, let me just explain what all those forecasts are. You can, they are not too many, so I don't need to write them again, I'm used to writing. So let's start with the number one one, one that we are all familiar with, this is a descriptive that we do a lot in business, right? So, management and the business folks, they tell you, if you're one of them, you really want to tell you that this is what you want to see on a daily or weekly or monthly basis, right? And typically, you want it to describe what is happening. So you want to see maybe our sales, how your sales are going to be by day, by some different kind of, maybe by food or by region, by sales person, right? So, there is a lot of descriptive analysis, in fact, some predictive things descriptive is like the role of what you can do in business. So that descriptive analysis, you are basically looking into the detail you have, going from download data to insightful reports, something that you can look at and see the kind of patterns you want to see, and begin to be one of the things that you want to be able to figure out, to do all the data here, which are being presented. Then there is the predictive analysis. So what most of us do in business is we do much of normal, you know how we call much of normal, then we do much of normal. So, maybe I'm a sales person, and then I get to work today, and my manager calls us or all these things and says, oh, we have an emergency meeting. So you know what, we had a meeting with manager last week, and a whole lot of things that have come up, your targets are going to be changed. And so maybe you used to do, say, 100 million per month, and they say, now your target is 150 million. And when you try and ask, how did they come up with that figure? It becomes the case of digital. You don't want to, you don't want to, you don't want to start working towards achieving, right? Magic number, you don't want to have to fall from somewhere. You can't even say, this was a calculation they need to come up with that this is the number you should do, right? So it's very common. MDC is going to do 20% here. Then everybody starts working. 20% came from, I don't know, maybe Tony Shaw. And then everybody else starts, you know, getting some other magic number that the few who have got the magic number, like Michael is an Indian, Michael, that's what he does, and I understand he wants to do his magic number. So that is one secret thing that he uses to do, like that's all he does that I do. Then the second one, we do moving average people. So we look at how you've been doing for like the last couple of years, periods, and if you do it 20% here or 20% month on month, you'll say, okay, if we ordinarily do this, I want to get to work harder. So if you're not going to get to 20%, maybe you do 30% on what basically comes down moving average, right? So those ones are very easy to do in nature. The magic number, you don't need to do any work, just ask the person giving the magic number, the right thing to do, right? And then when it comes to moving average, the last couple of periods, what about the much more intense ones? The ones that people, while you're in a zero data analysis, you can't do it in a zero, this come and then fight on, you learn to have. What about those ones? Are you now, let it help you when it comes to that and you're like, okay, if it's that, I'm not going to answer, I hope I'm going to do that. No, if you're an Excel person, maybe you don't have to go there, where do you learn something for, why are you learning those things? Also, you need to do that. There's something already very important that you can take advantage of in the Excel, right? So that's prescriptive. And someone said prescriptive. So prescriptive, that's why everything works, because it's when you combine with it. So maybe now, I put in a predictive analysis that can predict when, say, a particular product is going to go up to stop. So maybe we have to wear out, we have to have some logistics. And so that includes prediction. And then you want it so that anytime the product is going to go up to stop soon, it makes generating requests for output and sends to all of your mentors, right? It's just the information part. There's going to be more information outside. And there may be a feedback if you want to. Once you put it in, it makes it easier to understand. And then there's exploration. The exploratory is much more like what do you do to make a line, do what you want to do. And then most of the time I have a friend that does what he's into, when he starts talking about what he does in medicine. Like, are you doing medicine or are you doing, like, programming? So they call it, you know, there are some things we used to call that are just like... Aha, those kind of things are... What they are doing, their mind is black. When you're doing this, please, when you're doing this script, if you have an idea of what it includes, right? Exploiting everything, you don't have anything in mind. Just look at all the details, look at the patterns. So it's not really, it's not in flow stream. So the kind of things that you remember in the script are really like flow stream. Start with a narrative about some things that do market, market, basket, analysis, and segmentation. And it begins to look at that as a bit of kind of exploration of trying to make sense of things. Okay, so that's the thing that we're going to talk about. We're going to do things like... We're going to predict the whole situation, to keep people going, and then you want to run things through integration. All those things you mentioned when you talk about details and you have a lot of serious things that can't find in common, that's in Excel. Then the second aspect is the Excel JavaScript hat. How many of you are not familiar with it? You're very familiar with it. So we know that the people who are who, who's ready to make things inside most of the things, like automated communication, from scratch, from A, starts to make. And so you do the program, the whole thing, what you create, it's already transferred to the program and then why do you write the program? So I do this a lot to automate your photo, where it takes hours. But then there's something new, JavaScript. So this one is the whole point, when we get there, I'll explain it. So let's start with the first one. So, if you can see what I'm talking about on screen, let me explain what's going on. Say you work for a micro-finance, or you're giving back, right? You are interested in giving those regular types, right? That's why I said micro-finance, not feedbacks. So, if you now want to put some level of automation, do more with your data, right? So you have people who are taking notes from you before, and then there are basic information about them and it feels like useful. So their names, their gender, their age, their marital status, their employment status, their position, where they live, then their monthly salary rate, and then the loan amount it to, and then whether the pensioner needs to be paid for it also. This is information that is valid, that if you have someone who can do predictive analysis, you can look through it and find the pattern. And there are some people who, when you prefer them, so when you prefer them, when you look at their own parameters, they kind of fit into the kind of people that are involved, or the people that you are interested in, right? So how can I distinguish them? Well, I'm going to do an algorithm that I'm going to use a classification because it's two of the kind of specifications that I created before. So that's the binary thing to do. So we use something like a binary classification algorithm, and then you want to come up with the ability for materials to approach you, ask for your details, and I'm putting into the Excel stuff we're going to view, and it's a predict whether you are likely to be involved. I'm going to put down a vote, except for me. But the look of the idea would be to, first of all, define what factors, in the whole factor of determining something to go left or right. And for one, the way that I saw the list, how can I say unemployed people? But then, if you look at the multi-siders, how do people not take any above the middle? So there are things like investments, TVs, all that way that they make money. So if you can't just say no reason. So you have to, sometimes, maybe you have to pick some factors in between the two of them. And also, besides doing it fully, yes or no, can you complicate it a little bit more by saying certain things? Like, if somebody is unemployed, you, yeah, certainly this is like that. You can have a total, and they can complicate it. So the total comes... You are really the algorithm yourself. No, no, that's not my problem. Because, yes, it's supposed to use something that you're not the one who has no bias from you. You are the one who knows that I'm not the person of knowledge who's not the one who has no bias. But then, no, you don't have bias. You're the one who's exploring it. Okay, I'll finish first. Yes, sir. Maybe if you had units, if you had bombards somewhere behind the engine, attached to any of these components, you can do, you can go see, you can see where that number is and go for the new person's knowledge. So if I say, anybody, if I do this, who finds it, anybody who has less than 50 minutes will be found. So I'll probably just now set a goal-seek program at the end and I'll say if you just look for it, guys, who has more than 50 minutes. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Good. Very good. Good. Good. Good. Good. Good. Good. Good. Good. Thank you. Good. If you know statistics, you can always do this from post-printing. And you know post-printing means it doesn't matter what you have, right? You can't complete things out yourself. You have to clean out. We have to use the example you used something like this for. What we need to do is we need a regression, it's about regression. But then regression is not the best for a binary outcome. So we say, okay, let's try our example to be aggressive. Small problems are always going to be there, but then there's a lot to it. So we are trying to stretch ourselves. And we're writing the formulas and we're like using that here. It was a whole lot of work that we wanted to do. When I did it, because at the point we switched to hard, when we did it in hard, we were like, we wanted to do nothing because we were stressing it out. Because it's just the line code you put in the data, the data frame, and just like one. All of us spent like one day doing it ourselves. In like five minutes. And that's what I wanted to show you. How you two can leverage this new thing around the country. So I'm saying it's impossible to do this well. I also think you will ask me, am I already sure about that? You're in exile. Who stayed in the state of immigration or random threats? People are in exile. Some of us are in exile for doing some of these things. It's quite a lot of things. So we are going to find out. Are you ready? I'm going to do it there. Here I am. I'm working with this one. Because I need to have some... Okay. So I'm going to come back from now. I'm going to say it goes over there. So this is the raw data. It's about 5,000 records. What we want to do is... So what I want to do is... I want to do something that... I think it's the effects of data science. If you give someone who is a statistician or a data scientist what is they supposed to do? To build up something that we learn from the data we look at. So can you give them the activity or what's the difference? Are you saying 3,000 bits? Those that we pay, what are those unique characteristics, right? And then we waste them. So there might be some characteristics that are kind of like non-significant. You put the code and that's where last you put the code. Then you might be lucky that out of the 6, 7 characteristics there might be 3 or 2 that kind of you see it by actually you put it in the bit that you pay. And people maybe like to beat up the code, right? So you want your algorithm to do that. And you want to be... You want an algorithm that can keep relating to the code. You get more data maybe every 3 months because it's changed. Like there's no power names, right? So that's all we want to achieve. But then how do we achieve it? Are you ready? So there will be a small gtp. At the end of the... How's yours? That is the body or it's an embodiment of all Microsoft cloud solutions. So meaning that anything we want to do that we want to be able to leverage instead of building things ourselves. We want to leverage existing tools and in the minutes or later communicate with every single one of them which is what we're going to do. That's what we're going to be now used to achieve what I want to show you right now. So what I want to show you right now is... So this is what we want to do. So what I have is... So I have some random... I don't ask anyone to volunteer even though they're doing most of it before I sign in to volunteer. So I have some random... The issue that I have here now I've got all of their names. So maybe you might get that out of the phone. Or if you even integrate these with a web platform or a Microsoft platform. So if you integrate with that so that you can perform and that the back end begins to uplift and exert. We are named gender age. All this information that you can ask them that. But then, because you understand that we are trying to do your first command, but they don't know that if you are going to use it to take something, you know, more meaningful decision modules are being required. So I now have the amount you want to request for. I need something to predict. At least the express level verification. So almost can't take an intervention at all. But let's not run it through the machine, right? Let it tell us if these guys are doing things or not. So I'm going to select the data that I have about them. I'm going to... I'll show you how it's going to be. Let me just demo for you how it works. So I pick that and say yes. And then I need you to put the outcome here for me. And so I pick where I want the outcome to be. It's edited, right? You can still see it, I guess. That was that one. We'll make a note of the situation we have. So I wanted to be there so I can dance. Let me know. And what I will do next is I'll take you to make the pitch. And you're going to see some vocals, I think, right? And it's time for you to go dance. You'll see that it's good to tell me that these people are liking to be both. And this is the probability score, right? So I know that this person's probability of defaulting is a lot higher than that's how binary specifications work. The work on the A is the probability scale. So two things will be at the extreme. Probably if you're emailing and I want to ask some questions. So I could say it may be zero. It may be one. The probability is closed up to zero. And it's closed up to zero. So you see that everything is 0.5 and a boom will be tuned into another point on the scale. And then so that's why I know that this one that is really low is not closed up to that slide. So in case you... I mentioned that you look like a probability So it's the same thing that all of them do. And so how do you make it happen? You see that it's possible to have your central approach from the raw data to the final result. So we have to go back to the other move. Our zoom, since I will use the web in lock points, I'll be able to zoom in. You'll never wear this in contact with that zoom whatever you want. So it's product we are going to use. I thought it was Azure, right? But if you want to try it out, it's azurestudio.azureml.net. It's a big product. Studio.azureml.net. When you have done, it gives you the... So you give the algorithm in the end, it gives you any type of internet that you can use. So that's why I put that feature in there. I'm going to increase the zoom size. So the way to try it out, it's this studio.azureml.net. It's pre-account, everything is free. I don't think there's anything to do with this. Studio.azureml.net. So once you go there, you go in, you know, I don't like to stop the screen. It doesn't mean you have to, like, create a new image. And in the same way, when you buy a new tool, then you have to go and use any of it. So this one is in the way. And for me, because I'm logged in, I'll just go straight to my experimental best practices. So once you've logged in, this is what you will also see in your own case. So once you sign up and you log in, because I'm already used to... You know how close you are, if you want to experience it, this is where you see the menu on your left and you log out on your right. So it's not useful for us, because we have to carry it. So when you log in, you can just see if I can make the screen clear. This looks better. This looks better, right? So I can go ahead and insert this. So this is how it works. The data that I want you to make the prediction, so you wrote data that I want you to do the algorithm, because you must build up the algorithm first, right? I already have the data. This is it. So you would have given me your data and then I would have it already like this. This is the way you see it before. Whatever format you need, right? So this is what I will use to train the model, to build the algorithm. And then this is how, once I have that, I go to this place. I say I want to create an experiment. So I want to create an experiment. I do new. Can you see the new? This is that new. So once I click on that new, it will say I want to create, they have a lot of samples that you can use to get from it. So you need to see something like, oh, okay. You can analyze it, you can do all kinds of predictions. So, I'm not going to say about the experiment, but we'll do a blank experiment. I'm going to click a blank. So that way I create some structure. So I click on the blank. It gives it a generic name, right? So I'm going to give it a name that I feel is very much what I want to do. I want to do a slow prediction. So once you give it a name, you don't have to do anything of the same, right? And the first thing that we want to do is, I want to bring in the data that I have that I want to use to. You can do analysis without data. So all the data you say, you can't do this, you can't do that. And then you forget about it, you can't do that, you can't do that. So, I already have something in data. It's very easy to bring in just the important thing. So I have it as my data set. Let me see if I can show you that state. It doesn't look like that. On this left side, experiment, there's projects, there's where they are, there's some things. The one that I want to talk about is the data set, because that's where you need to start. So I want to bring in the data set. The new is here. There is a new lower left. You see the new here, right? So that's where I'm going. And then I see that I have it in my computer and I go, it's here to the file. It's like that. And I say, yes, bring it in. So that's what we're doing. So I've already done that. So that's what I'm supposed to do. So once you do that, you will see it's here. So I'm just going to take what I have. To see if it's a drag and drop. So, you're not going to be typing anything to your... Everybody knows how to do that. They don't have to go like that. So, I'm driving the data set, right? So I think at this point I can go back for everything I need to do. So you're going to find some things that are similar. Why not bring in data? When you want to do a 3D algorithm, you want to separate the data into two sets, a 16 set, right? Doesn't make sense to ignore your data with an algorithm. How do you not test the algorithm? So you split it maybe like maybe 64. Use 63. I'm going to use 42. So that's what I'm going to do. So I'm going to say I want to split it. I want to split the data. So I need to split it. I drag this there too. And then I connect it to... I connect this to this. Then I say I should split it in there. I'll rather... Instead of doing that, maybe you want to use 64. So it's always the left. That's your control. So that left, whatever you put here, you want to make sure of the data around your left side. And then I am going to now say this data... By the way, there's something I would like to do here. You know we have your names. I don't know if you remember. We don't have names in the prediction, right? So I'm going to tell the model that I don't need any names in the prediction. I'll read your name. Whether you feel or not, please. So I want to select some of the... I'm going to say I don't want to use all of the data. I want to select... It's not faster if I just write what I want to do. It has to... it has to be something. So once you say this is what you want to do, type something that you don't have, and I select columns. So I'm going to do that. In fact, I prefer to do that before you ask me. So when I select a column, I get this column, column selector, pick it and say the columns I want to include. So you see, I've probably got the columns. Because I want to do it faster. I'm not letting me do it really fast. Let me just briefly show you what it will look like on my notes tonight. Because the doing it is just dragging. And the first time I showed you what to be dragged. But I will focus more on the aspect because I still have like just... I'll show you what to do, all the things that I am planning to drag there and then the end results. So this is... This is where we... the data, right? The select, where I'm going to select the columns that I need to use. So this is what I was trying to show you. So I split the columns I want. I want them to be back home by one minute. Yeah, I want them to be back home by one minute. So yeah, I take out the name. Right? And then I split the data now. So you only call them two columns. You don't want to take it in the name. That's all I say. I don't need your names. All the other columns are there. All the other columns in the data set are there. Yeah, except for me. So I don't think I need a split. So I do this split. Which is... I'm going to use the answer for the descent there, right? Then it splits into... I split it into some... some more... Yeah, except for that one. So I split it into the left. It's going to be my trading set. And it's going to be right. It's going to be my trading set. And then I really... I need to... I don't know if you can see it. I think it's going to be the trading set. It's going to be more there. And it's going to be the trading set. Where I connect the output. I need to... I was very fast on the back. So that's what I'm going to do. So I'm starting from... I'm dragging the data set. I want to select... the columns I need. So I click... I click for the select columns. I connect the two. Then I want the column selects. Then... So you see, it displays everything inside. So you see, it displays everything inside. I want everything except... the name. So you see I pick everything except the name. I move them to the other side, right? Then that's what I want to do it on. Then I'm going to... bring in the split. So on this time, you will see that I have 5 minutes to come. That's why I click this column. I connect the split. Right? And then I say... I want it to be less than 16 columns. Right? Then I bring in the trade model. Right? So it's everything I type in. It comes up with the... So I'm bringing the trade model. I drag that right in. So as we are dragging it in, it needs to start... to make it... So the trade model I'm going to... do the... connect the data here to this. Then I need to... use an algorithm for the trade. So I do that... to find out... that I like a classification... a... some cloud can classify... to one of the... So this one... I don't type... so I can see all the options that I have. And you know... So if this is a machine... and then... a different anomaly... classification... is what I want. You... I always have... this thing called regression. So one of the things... it... it does it very well with this kind of data. So I'm going to drag it into this regression. It's like regression... but... so I drag that in. Then I'm going to... evaluate the model. So... evaluate the model. I bring that in. I connect the... output of the model... to the evaluate model. I connect the output of my... spring data to the other side... that I will have to test it, right? And then... So once you... this is what I'm drawing. You can see it like this. So once you do this... once you run it... it will allow you to deploy it. So is this possible to expand... the model? Okay, so like this. I'm using the idea... when you know it's let go of... and it's going to run... and then it's evaluated... and then it's run... and then it's out. So once that is done... I run it... this is run... and after I run it... I will deploy it. So... I will deploy it... after I run it... I'll show you. So... So if you connect this together and blah blah... then you run the model. You run the model... then you deploy it. So if you run the model... then by running it... you... at what point do you run it? At what time do you connect? All of the time. So once I run it... once it runs successfully... it may now be the option of... set of... I need to set up web service. I need to... now... So once I did the set up... set up web service... and I launch it... launch it in the other part... which you will now be able to exert... as you will see... a deployment... Once you run it... you will get... I need to deploy... after I run it... that is what completely might... once you can deploy... once you can deploy... once you can deploy... deploy... So it's trying to deploy... I can quickly switch... when I run it... I can switch to the complete part of mine. Once that's the last bit... you will see an example... that you can... So that's the predictive side... and everything... has all the different types of predictive... are going in that way. So that's the predictive side. I think you know in Portugal... I'm sure you know how to do it... not as... I just want to also make sure that... I show you the next part too. The next part of the presentation... which is... Excel, JavaScript... I asked us that... we are familiar with DBA, right? And then... So this is the one... I don't know... The other one... is the new hardy experience. So this one... those of us that are... who might be enjoying this one... because... one of the problems you have with... you have with the... existing DBA kind of adding... is the fact that... you... it can be exploited... I think some companies are... they disagree with something like this. So if you try to put in any program... that has a DBA... it's not going to be wrong, right? So what are the ways you can deploy things... and use without you having to... run the DBA? You can use in Office 36 Friday. Thank you. If you use Office 36 Friday... if you go to INSAN... you will see hardys... you use Office... Office 2013 and 2016... because... if you use Office 365... you get a lot more... and so forth... you can explore. So if I do this get hardys... you will see it just can be right in store... using the old versions... or if you use the old version... you see get hardys... it's going to show you... the whole model of the tools... that you can use... beyond what is needed in the Excel, right? And then... if you scroll down... you can go into... you see some that are useful for you... some might not be... so now... are you seeing something familiar? Can you see something in Nigerian market data? I think that I need to put it up... so you can do something too... if it's not just your organization... if you want to do something... anybody can install, right? And so like in my case... what I built is something that I can allow people to use... and in some other cases... you solve the problem of distributing you all... aha! they can even pay you... you can only price three minutes. If you have the problem of distributing without something... then you have to start doing... paying for marketing. You can't just see it without me having to... give them the need to... and then if you have... it's time to add it to your Excel, right? And then... so what I've done with this... what I've done with this... I made it possible for you to... do... information that we all like to get from Nigeria... like you want to know what the effects are... and you also want to know what the parameters are... and what the effects are... one of the commonest things you keep telling me about... and that's the aspect you want to know. So I go there... I say I want to see the... exchange rate... and it's time to... I want to see... maybe the stock market... I want to see the stock market prices... and it's time to... stop and choose me. And so today... it should be equal to 29... because that's the last market... right? What if I want to know what it was... this same year or last year... right? If I type in... if I type in this same year or last year... 2018... I think 6... I think 79... and I take it... and I say to get before that say... things... and I say to myself... and I say to myself... and I say to myself... and I say to myself... let's even go 2 years back... less than 27... I like to see... and if you can... so this can be for those who have investment portfolios... and they can go back in time... and the same thing... for the exchange rates... you know... you guys forgot to put the closing... exchange rates... I can't predict... what about 2019... 2003... end of the month... whatever it is... I want to get what the exchange rate was... last year... and I think you have particular things... some of the elected things... and you get it out for you... and if you like to get... this is the one I put there before... you say... exchange rates... so you have a significant rate... and I raise it... so all those things are there... but the thing I want to mention... is how do you use something like this... because the key thing is... you want to make it... see, I didn't have to install the MP... you can use it on your phone... because the same functionality is on your phone... and then even when... IT puts the password... to give you all these... and give you that right time... and so on and so forth... because you actually know it's going to be... there are only 18 or something like that... so how do you do this... is... it's going to be part way... some people might be like... okay... and now I know it... but I will give it to some other guys... you will need some help... yeah... you will need to have bigger skills... and then you will build it inside it... what means you know it's impossible... so if you have a very large organization... and you work in IT side... this is something you can learn with that... so how do you do it... if you go here... so now we are getting to the question time... if you go here... the same place I want to do the... get add-ins, right? you will see my organization... so there is a place where... if I could... there is a place called My Harding... there is a new manager... a new manager is... an organization you can publish... things that are just where you are... in your company... so now we are on our way to deploy things... things like that exchange rate... so I have come to the end of my presentation... unfortunately do you want that... I was really wanting to know... how you are taking... okay it's finally done... deploy... so you say if I deploy... and then I want to deploy... any questions... so I can keep the question on there... so when the deployment is done... can you see something like... let me just close all this... can you see that I can download... the actual part? this is magical... this is what I now can do... that now makes this... like we open it... you now see the same experience... you saw me use to answer... and then I can type in any questions... oh I need to... so this was the thing... so you now see... that you saw me... using the demonstration... that's the only bit that is outside of it... the moment you do that... everything you see... right here... this is how you make it... any questions? I saw you scroll... scroll more... more positive... the... in picking the correct model... or whenever you go in the model... you want to know how it performs... and you got the... testing... that means right... so that's what that's for model stores... for those of us... I'm actually not showing you the results... so this is how we have... we have... we have... so this one we now give you... there's a technical way of doing it... you call it... true positive... or future measures... so it's very easy to see... true positive, false positive... and it's called the accuracy and saving... it's like you're waiting to use the testing... so that's what I think I want to be showing you... so it's a long question... you're a pressure model... sorry pressure model... how did you... how did you... how did you... how did you... very good question... remember I said that everything can be done from first place... this question is the... the regression model I gave... what did I... did I set the picture... did I... good thing with... of the shell... product is... they can... I think... find the best setting... I don't know if you have used forecasting... if you go to data... maybe... you will see something called forecasting... how many of you have used it? forecasting says that... it's a time series tool... but if you... if you don't know the strategy behind you... you cannot take it as trivial... what it does is what we call... people... explanations... like... find seasonality... find strength... find the error factor... that thing I've said... if you try to do it... you know, outside of Excel... you can... but they put it there, right? and the thing is... because... they don't expect you to restart the shell... you can automatically determine some analysis... which is the one you can do all those things... you can find the best setting by itself... because it can also score itself... and it won't be... so that's... I don't... I don't... I'm doing a competition... and I want to do it like that... so... so if you... if you want to... if you want to... if you want to... if you want to... so... if you use this thing we have done... it's just the guitar set... to replace... and those we call... so that you can have a drink... enjoy... you will know... you don't even need to re-download this game... you just put it down... replace the guitar mode there... play it for a bit... and then have it... you can keep the game... so... I just want to confirm that you don't have a tax code... right? around this big data set... because there are many expectations... about how common you are... so... I don't know if you are... if there is already a significant population... it's a significant sum... if you use this formula... it's almost like a rule of thumb... the formula to use to that... it's easy to find out... the population of... 2 million... what sum to size can I take as representation? so that's what people have even told... the direction and government... and all these other rules... so they know that this is the target nobody needs... for the customer... so... the first thing is... once I know that... amount... I don't know what my data set is... but I have a data set of... based on... if all our customers... if the... what you can Google... but you can Google it... something like... if people have few tools around you... say... so if they can say... something like 2000... if you give me... if I have the computer that I want... I will have done that... and say... because people can talk now... you're like... you just give me my computer that's good... so I will now use that... back to how it comes to your question... so now when I know I have 5000... but if this is I need 2000... then in that case... I can play... I don't have a lot I put on the testing side... it's more than 2000... so there's no more... or if I have 3000... I'm most... start to... start to size requirement... so that's where it's more... just you come to the step... so you select data set... select from the split data... bringing data set... I select... because sometimes it's very common... that there would be some... maybe a confidential... or things that... that's what you select... split data... then... you start to do two parts... the logistic... you change on frame... but... is that even what you use? the train... it's called... so the way it works is... that one goes to the train... the train goes to the score... so that's called... the train goes to the score... and the second side of the other... goes to the other side of the score... and now... the scores will go down... so that one goes to the evaluation... so that... so how do I evaluate to run? and so on... how do I evaluate to run? that evaluation... what part do you run? can I click on evaluation and view... without the accuracy? so I'm going to click on that... show me that... I'm very confident... but you cannot read it... 60% actually... so sometimes I think there are cases... I'm going to change... I'm going to do my evaluation... but after that... I'm going to do my evaluation... so find out... how do you... so how do you find... how do you find... how do you find... how do you find... how do you... read it or say it... how does that work? is it something you can't read... or it's already set up... or it's sort of thinking that... and I think I have found it better... with the evaluation... you know I wrote it in detail... I'm not sure if you can... but that would be better... so you can let it adjust... but you don't want it to be... wanting to... take the point... the only thing is... you might need to... you know I'm good about detail... maybe you should bring the last 5,000... so you know how you need to... select the top... one column... so without... what is that you don't like asking questions? I like... okay I think that's enough... let's talk from... how much of... how much of the... theory of satanic... you know... which is that... say... like now you know your name... binary... so when do you know... how do you know when to use the... what component are you using... and how do you... you know I'm just saying... how do you know... how do you know... how much of the theory do you need to know... before you go... and my time is up... so I'll just take the steps... so I don't know either... the geography... how does that make it... so if you have a data set... a lot of data sets probably explains... the americans set for example... if you bring it to... the migrant sector... if there is anything you need to do... to account for behavior... that what you get as religion... kind of matches what happens... you don't just take a test... it's quite good... so if you... can I just go on and pick a model... and say this is how it is... and then three months down the line... you know... let me answer the questions again... so question number one is... how do you know which app are you going to use... how do you know how you feel about those... do you know how to include these types of things... to things... if this is a part of your life... you want to be a professional in that part... then you can buy a part of it... on that account... or if this is going to be a one of things... maybe something... not like maybe you want to switch career... then you can just read... actually three bits of what you want to do... I can do that... in a packet packet... packet packet analysis... or something like that... just move on... just focus on the things I want to do... and use that language... so that's what I think... when I have things that... I know what I want to do... then question number two is... can you take something... that someone has used outside of the middle... to read... okay... the good thing is data is not... we are going to know that this is injected... data doesn't know that... so the only thing is that... this time I will be different... so it's more about data itself... so it's always good that... algorithm might be similar... but it's good to run your... from scratch... you might use that as a guide... use that to do your first run... then start fine tuning... but it's always never... correct your thinking about... about human life... you will always have to do your fine tuning... even in the same manner... to defend confidence... thank you very very much...