STATURDAY CHAT LIVE with Brandon Foltz, Ep. 62

n n hello everybody and welcome to 2024 this is the first live stream of 2024 of course I hope you had a happy New Year wherever you are in the world and the new year is always an opportunity for reset renewal and reinvigoration and that's what I'm doing here in 2024 so I hope you are doing well also so if you're new to the live stream this is kind of how this works I usually go 30 minutes to an hour depending on how the conversation goes the chat is for you so if you have any comments questions want to get some discussion going that's the place to do it so go ahead and put anything in the chat if you want to get the ball rolling and we will put your comments or I up on screen and we will get going so I'm not always able to answer every question here live in person so if that's the case you are more than happy to contact me either through Linkedin which is probably the easiest way if you're on LinkedIn or through my website which I'll show you here in a second how you can do that so yeah so I always like to know where everyone in the world is watching from so uh please let me know where you're watching from in the world and we'll put that up on screen as well that always gives me potential travel destinations as I Venture across the globe here in this new era of my life so let me know where you're watching from um so let's talk about what the Channel's going what's going on with the channel and what you can expect going forward in 2024 so if you are a subscriber to the channel and I hope that you are that's why you're here you will see that last week I released the very first video in a brand new series of many videos that are going to come and it's based on uh it's called machine learning for M morals it is a way of exploring the concept of machine learning for individuals who are new to it who may have been intimidated by mathematics and coding and so on and so forth so it could be a university student that's just getting into it could be a high school student that's kind of advanced uh in math and wants to get into that or a professional that you an older professional like me I'm almost 50 years old at this point who has always wanted to learn about it but what felt kind of intimidated by it so that's who it's geared towards and we are using a program called Orange it's a free open- source piece of software called orange that runs on Windows and Mac OS so you can download that and follow along I and I want you to follow along that's the whole point of the series so uh new content this is something that many people on the channel and on this live stream have asked for years so I've gone ahead and decided to do it uh it's kind of was the time for me to go ahead and explore that again new year new ideas and you know I just wanted to explore machine learning for those individuals so that is the plan um going forward all right so if you see me kind of doing this a little bit that's because I just got finished exercising so I'm still a bit sweaty and hot from exercising so um I'm not like ill or anything I'm just sweaty so kind of warm from exercising so let me talk a bit more about the series that's coming up and sort of what you can expect in terms of how it's going to work and topics and so on and so forth so the next video I'm working on right now is sort of a way to get people to think about machine learning in a different sort of way and to get the mindset for the things we're

going to learn uh going forward so the first sort of playlist in this new series will cover some of the basic concepts of course so what machine learning is how we can think about it and then fundamental concepts like how the data is structured in many cases so features and targets things like confusion matrices area under the curve uh stuff like that so we're going to cover some basic concepts cross validation train test splits so on and so forth then we'll get into specific use cases like decision trees and other models and sort of how they work from at a level that anyone can sort of grasp so I'm not getting into the crazy math of anything it's really you know conceptual using orange to explore data and build model workflows and so on and so forth that's what it's going to be and I already have some of the future videos already sketched out um some basic scripts already done so that's going to be the gist of these uh next few videos and of course I do this I do YouTube when I have the time it's not my full-time job by any uh by any means I actually lose money uh doing YouTube so I'll get things out as fast as I can depending on my schedule but I am committed to making those videos uh for you and the rest of the world for that matter so you can look for those new videos uh coming up all right so that's what the channel uh is to let's see um uh Focus I assume I'm saying that correctly hello Brandon great to meet you in my undertaking you're making my journey to be easy and short God bless you so much thank you for that comment I appreciate it that's kind of why I do this from Tanzania that's amazing you're making my journey of studying to be smooth but you are very very very welcome that's why I'm doing it and of course the content on my channel is really geared towards individuals like you focus who just need a little bit extra uh support and help along your Learning Journey and for individuals who are new to these Concepts who may have been intimidated by them in the past like I said before and I try to explain them in a way that matches How I Learned them in my own head if that makes sense so these videos do take time because I put a lot of thought dozens of hours of thought um into them so you know people will see me just staring out into space sometimes and when I'm doing that that's me thinking how do I want to arrange the content in this video what examples do I want to use uh what features of the software do I want to use to explore this concept in a way that makes it understandable and approachable to everyone who watches it and that takes a ton of thought as any uh teacher will tell you they spend the vast majority of the time just thinking about the best way to approach this to individuals who might be learning it for the first time but I've been doing that for the past you know a couple months probably with this new content so I'm pretty good as far as the first several videos go and I look forward to getting those out over the next couple of weeks just depending on how things go and so on and so forth so that is what the channel is up to and we'll we we'll see how it goes but I'm excited to get that going all right okay so what other comments questions do you have go ahead and put those in the chat and you know we'll address those also want to sort of encourage you as I have been with the past several months on this channel is to really make sure you're staying on top of all the developments in AI it is like drinking from a fire hose you that sort of meme or proverb proverbial drinking from the fire hose because it changes so much but I think it is very important for us that work in and around data specifically to really stay on top of those uh changes and the evolution of these the AI models and so on and so forth so we can stay on the Vanguard of these new technologies over the next several years probably it maybe forever imp impetu going forward perpetuity going forward we will have to be constant Learners all the time because this stuff changes and evolves so fast that I can make your head snap so I would encourage everyone to keep doing that keep up with the news there are some great AI channels on YouTube that you can follow so if you start looking you know searching for that kind of stuff they will pop up and I can show you maybe a few of the ones that I like here in a minute um but always stay on top of these developing uh Technologies because this is the new reality this is what we're going to be living with so I encourage you to do that also it keeps you out in the front

in terms of not only your knowledge but also your employability because it will be very important for those of us who work in this spa space working with data teaching these Concepts or what have you to really stay out in front so when our companies have to evolve and our businesses have to evolve we are ready and we are looked at as the the resources to go to when people want to understand these sorts of things so we have to be experimental out on the edge and we have to be doing some trial and error and seeing how these things work so I encourage everyone and of course me I that's what I do all the time is stay on top of all this stuff like I said on this live stream before I work my day job from about 9:00 to 5: and I'll usually eat dinner and then I'll spend a good part of the evening after that just watching videos on AI uh doing research for my new videos on this channel and all of that it's almost like a second job after my actual job and I think that's just going to be you know the new reality going forward so I hope all of you kind of have that mindset too because that's what we're going to have to do you know that's kind of what we're going to have to that's the mindset we're going to have going forward if we have any hope of staying on top of this stuff that changes so rapidly okay um so speaking of a few things I will share with you let me share my screen here all right so obviously this is the channel uh if you haven't already make sure you do subscribe and hit that Bell notification I don't publish like every single day of course so having the Bell notification isn't going to drop drive you crazy I'm not going to be you know hitting your notifications all the time but because I release video videos on a fairly rare basis because of you it being a part-time job it's just good to have notifications up so you know when they go live because I don't go on a normal schedule like a set schedule so I know many people tell you on their videos so please you know subscribe and hit the Bell notification but in my case because I release videos relatively rarely and not on a set schedule it's good to have that Bell hit or pressed so you see it come up when I do publish okay in my case all right so obviously here's the channel you can see the new videos right here um I am kind of doing a little bit different with this one in most of my prior videos I haven't been really on camera because I did not want to take away from the content that I was teaching because I think it's kind of hard sometimes to see your you're trying to watch you know the lesson or the tutorial but then you have this face in the corner and you're kind of going back and forth between the two but I'm going to try to you know put my uh the face the of the a great face for radio as they say put that in the lower corner so it's a bit more personal you know for you all you can see me there so um unless it gets in the way but yeah so here's the first video you can see here and um I'm going to try to keep them relatively short and they will be how-to so like I said we're using a program called orange the whole idea of this is that you download orange install it on your computer and follow along so many Learners want Hands-On uh educational material they want Hands-On lessons well if you want Hands-On lessons you have to download stuff and do it um and orange is it takes two seconds to download and install and it's free and it's great so make sure you download orange and follow along with the videos as I go not every single video per se will be using orange some of them will be more traditional just discussion around certain Concepts that we can then apply when we get into orange okay so that's part of the reason why I made my new thumbnail orange uh for the live streams because every year I use a different background color so it was blue and then white and now it's um orange and since we're using orange a lot in this new series I just decided to make it orange as well plus I like orange so yeah make sure you download orange when you follow along with these okay I do I am going to set up a GitHub uh site where you can download data sets you can download the orange workflows and any other um assets you can follow along with so I'm working on putting all that together I don't really have much to do to put up there yet obviously because I've only done one video but I will have something where you can actually just download the files and follow along I might have the

workflows in Orange in different stages like a beginning and end like you would see maybe on a LinkedIn learning course or you to me or something like that and you can you know start and get to the end point all right and if you have any questions about orange I'm happy to answer them on the live stream it's a little bit easier to do here than via you know chat somewhere so I'm more than happy to answer that so obviously while I while I'm talking about orange this is the website so this is the program that we're going to use in the machine learning series again it's free it is very user friendly it's open source and you can download it for whatever system you are on I go over this in the video by the way so you can uh just download for Windows or Mac and then follow along okay they also have some great resources here if you want to explore orange further they have plenty of screenshots from the different workflows and you can see how they Implement different things this program is so freaking cool I just I love it love it then they have some examples so if you want to follow along you can actually download the workflow uh file that they have here and follow along okay so these are like little tutorials that they have that you can follow along with you download and so on and so forth um the blog they'll have different applications so you know for here so this goes over Orange's scoring sheet widget and talks about machine learning predictions using a simple scoring system so you can follow along and read the blog there and uh all the docs are here so if you're interested in their widgets you go the widget catalog over here and here are all the widgets that they have in Orange and what they do you can click on each one and it'll tell you sort of what they do and of course we'll be covering a lot of these uh widgets not all of them but many of them in our video series but if you want to learn some more on your own you can go here and um do that and of course they have all the well they have the docs that's where the widgets are so that's orange you know this is the program we're going to be using and I highly encourage you to download it and install it on your system or systems it's great for exploration it's great for um uh interactive type things and you'll see why as we go orange is not necessarily something you're going to build like a production level machine Learning System on but it's a great sort of blueprint to do that so you can take a smaller set of data and kind of run it through the workflow and kind of see how it works and then um you can maybe apply that to a bigger sort of thing but again my series and orange in particular are really geared towards Learners that's a strong suit so that's what we're going to be focusing on but it's great to think about a smaller project before doing something bigger and it does it runs on top of python so everything that orange is doing is just python it's just in a very visual you know userfriendly rapper okay so that is that um obviously on my website here you can contact me right here so if you have a question for the live stream you can always submit it here and again of course on LinkedIn you can get a hold of me uh as well okay so I'm pretty active on LinkedIn I don't have any other social media I deleted all that several years ago one of the best things I ever did um but you can contact me on LinkedIn or through my website okay but linkedin's usually better LinkedIn is usually a better way all right also Clyde hey Clyde good to see you again excited about the new series yeah I'm really H I'm really excited about it too you know I'm looking forward to uh I've been using orange for years and I've been wanting to do some machine learning content for years because many of you have asked for that uh for a long time and the two kind of came together right now for me to sort of want to do that and feel comfortable doing that and have time to do that so it all came together and I think people will really like it you know so but I'm biased I'm biased right so what else can I show you while I have everything up let me see here I think that's pretty much all I wanted to show you okay so just make sure you subscribe click that Bell notification so when I do release videos which is not you know on a like I said on a schedule or very often you hear about them okay I don't do this for the fame the glory or the money because there's none of this none of it comes from this I do it to help people so if you want to be helped by the content it helps to be notified when it's going to be released okay all right what other comments questions do you have I'm gonna take a drink because

I've been talking for about 20 minutes what other comments question thoughts do you have it can be about anything regarded to the channel some of the content we cover on the channel it can be current events um related to what we do or what have you so the floor is yours while I take a that's my post exercise smoothie so I'm waiting to see if anyone has any comments or questions uh I'll kind of tell you the live stream schedule coming up so as of right now we'll be here in two weeks which is the 27th I do have some family stuff going on right now so that might change uh depending uh you know on those that those circumstances but as of right now the 27th of January will be our next live stream which is two weeks you know two weeks from now right um so Karan Karen I hope I one of those two is probably correctly probably correct I'm sorry if I mispronounce it thanks for everything you've been doing you are very welcome it is my pleasure to be there to help you all and it's fun for me too because you know what you see on screen when you look at my videos on YouTube what you're really seeing is a manifestation of the thought process in my head about one how I learn these things to how I break them apart into pieces and kind of make sense of them and then put them back together to help you learn so really my videos are this kind of weird physical manifestation of my brain and for me that's really fun to do so the idea of taking these Concepts and really thinking how do I want to explain them how can I get these points across in a way that is accurate informative and helpful and maybe entertaining so that you all can benefit from them I'm not an Entertainer so my personality in real life is actually pretty crazy those of people who know me but in the educational content I try to stay you know sort of centered and not go you know Bonkers off the edges so I really want the content to be helpful and accurate and informative okay so that's my style but you're seeing it on screen right so I got yeah I got one of those right appreciate that um Clyde says on UNC upcoming project I might be working with a data set from an LMS so for those of you who do not know an LMS is a learning management system popular ones would include canvas Blackboard D2L Sakai was one I'm not sure if Sakai is still a thing so there are that's what LMS is so if you work if you go to a university there's almost a 100% chance that you have to use one of those lmss that's what LMS is to analyze how students are using the learning materials do you have any advice or insights that might help maybe a few so actually may maybe more than a few one thing about one thing I will say is depends on the number of students you're going to work with Clyde so a class size of 20 to 30 is a pretty small sample right so when doing your analysis proc with a smaller sample size you have to keep in mind that statistically speaking making inferences from that small of a sample size is going to be fraught with problems technically what that means is that your confidence interal is going to be a mile wide and that's if you're using the whole sample so making inferences from a smaller sample is one challenge you will have to face so you might want to employ some non-parametric techniques if you're using that smaller sample size and I have a an entire series of course on non-parametric methods if you want to make inferences that's the first thing now if you have more students if you're looking across multiple sections and the student count is getting into the hundreds then it becomes a bit easier to make inferences based on a sample that size if you look at it as a whole also keep in mind if you're going

to look at subgroups of that data cide that you also have to have enough uh individuals enough students in that sample to make inferences so you have to think of the smallest possible slice of the data you want to make inferences about and realize you may have some limitations in the sample size when you make those inferences due to the nature of uncertainty and confidence intervals uh for that matter all right so again if you look at the margin of error you know formula the denominator is a square root of the sample size and that's where that comes from so just keep that in mind I will also say that depending on the data you get from the LMS it would be of some benefit if you have some more some qualitative data as well that you can use uh to do that sort of analysis and then keep in mind the way that the the learning materials are inset in the LMS so you know that if you're if you're counting like how many students downloaded this or how many students accessed this piece of content um how long were they working with this piece of content um you know what was their scores and so on and so forth keep in mind the limitations of the LMS itself in terms of how the data is put in there and what you're actually getting you know from it because students interact with online systems and very strange ways that you wouldn't necessarily anticipate and you have to keep that in mind too because you're not there watching them actually do it you know you could if the LMS as a student spent one hour on a certain activity it could mean that they spent 10 minutes on the activity and then were playing with their phone the other 50 minutes even though they were logged in to the LMS doing the activity so you have to keep like user behavior in mind as well but you can do some interesting visualizations with that data again depending on how many students you're looking at um you know and U if you're working across several sections keep in mind the confounding factor of different instructors as well so you know there it again I don't know the data you're working with I don't know how it's structured scope or anything but there are different confounding factors in data like that as in all data that you have to keep in mind right so class time you know if our class is at 8 am then I'm talking about campus you know or synchronous um class a class at 8 a might have different behaviors than one that's at two in the afternoon you know so you have to keep that in mind too but just you know think about the problem in multiple for multiple facets multiple dimensions and realize that the data is embedded in a certain system literally and figuratively when you go to analyze it so those are couple of um thoughts that I have on that but I'm happy to elaborate if you have if you want any more um more questions or elaborations on what I said roav hello Bangalore excellent love my viewers in Indie of course this channel would not be what it is without my viewers in India and increasingly uh in Nigeria um I can see through my analytics that that's picking up as well so I'm really excited about that but it's good to see you rakab I'm glad you are here thank you for stopping by you know and the and on this sort of idea of data and kind of going back to Orange is that there's so many different kinds of data you can work with easily in Orange you can use regular tabular data like we're used to but you can also use images there's image analytics in there is textual analytics text um in there uh you can do all kinds of crazy cool stuff you can actually connect to databases using it so we willing to experiment with whatever type of um content you have and you can get in there and play around it's really really cool all right whatever questions comments thoughts do you have while I take another drink and I will say also if you watch the new content and have any comments or

questions or suggestions this is definitely sort of a because it's a new type of content a new series of content I'm kind of developing it as I go it's not like I've been doing this you know this by the time I got to the end of the stats stuff which I will probably do some more of at some point I kind of knew like what I was doing and I kind of had like a set way of doing them with the new machine learning stuff I don't really have a set way necessarily so if you have any comments or suggestions you know towards the beginning is the best time to provide them and I will definitely take those into account sort of as they go forward okay so I always welcome your feedback as well all right so I'll wait another minute or two for any comments or questions and then if we're done we'll go ahead and wrap this up so this is last chance or forever hold your peace and then I'm going to get on with my day I'm going to go up to the uh out and start working on the next video so I have I'm doing a bit more scripting with this series as of right now to make sure I get everything down I want to say so I usually don't do scripted videos you know partly I didn't need to because the PowerPoint I was using was sort of like the outline of what I wanted to say but in this series because I'm using orange and like doing stuff I got to have sort of a loose script or idea so like today I'm going to go up and work on this sort of the script for the next video that's what I'm going to do when I leave this so um I don't see any other comments or questions so I'm not going to keep you all here I can let you all get on with your day wherever you happen to be but I'm excited that you were here and I'm excited about this new video series that I'm working on it's really exciting for me personally to teach this sort of stuff in my way that's unique to me and um yeah we'll see how it goes but so I'm going to go ahead and sign off for now and head up and do some work so thanks for stopping by and happy 2024 to you and I'll see you again probably in two weeks okay all right take care bye