Udacity AI Nanodegree Project 4 Planning and Progress | Learning Intelligence 12

What's going on, learners? Welcome to Learning Intelligence Episode 12, and today's been a great day. I've got a fair bit of study done. We look at my Trello board. By the way, you can access this. I'll put it in the description. I've got all my curriculum here from my self-created AI master's degree, putting together online courses to learn artificial intelligence. So what do we do today? We can tick off this. We reviewed the Hidden Markov Models class in the artificial intelligence category, so we'll move that to the done. And we also finished some readings in the textbook, so required textbook readings. I'm a bit behind on that. Weeding's— I should say readings, ah, not weeding. So, that's going to go into the done box. If you're wondering what sort of method this is, I have a Kanban method. So, I have resources, projects I need to do, and the last project is actually in the doing box. But things I need to do, I put into the to-do column, and then things I'm doing, I put in the doing column, and then I've got here like monthly done, so I can sort of keep track of what I've done over time. As you can see, this goes right back to March at the start of the year. What else have I been up to? Well, I watched a really awesome video this morning. And I'll put the link in the description. It was on dual process theory, and which is from the book Thinking Fast and Slow by Daniel Kahneman. Apart from the side that I really like his first name, the book is actually incredible. It's so in-depth and he's got a, he's a famous psychologist who's got a Nobel Prize in economics. I don't know the full backstory from that, but essentially what the book is, is goes through two frames of mind when the mind is thinking. So there's number, there's state one, which is the fast thinking frame. When, for example, I walk around the room, I'm not thinking of moving my legs. Just move for me. But if I go to solve a coding challenge or a math problem, I'm using the slow part of my brain, the logic and reasoning, whereas the intuition part, that's the fast part. And what this, what this video went through on dual process theory by, I believe Monica Anderson is her name, is a new approach to studying AI. And it's sort of, some of the points in there were mind-blowing. And her argument is that we've been studying artificial intelligence wrong. We've been focusing too much on logical reasoning rather than intuitive understanding. And I'll put a link to the video in the description so you can check it out too. And I'll let you draw your own conclusions. If you watch it and you think it's interesting, leave a comment below or leave a comment on her video so that it can get out to more people because I think she has a— she has a pretty valid point. Ayo, Nana Gree times 6 Pomodoros, check. And now we're gonna do a vlog clip. Well, that's what we're doing right now. This is like a list I set for myself every day. Put the date up here. I try to keep it to like 6 tasks. Sometimes it just gets out of hand. So I got the noise-canceling headphones on, I've been programming, had the Conquer playlist on. This is in iTunes. It's just like a whole bunch of tracks that really put me in the zone. It's sort of like Brain.fm, but my trial membership ran out on Brain.fm, so I haven't decided whether I'm gonna sign up for the premium one. We're working through Project 4 of the Udacity Artificial Intelligence Nanodegree Term 1. Actually, today hasn't been much coding. It's been all about planning. And check this out, I got pages and pages and pages. Here's where we started. So we got 6 pages worth of handwriting notes, and then this whole document is devoted to Project for planning. So I've got all checklists here. And if I didn't have enough checklists, I've got another one on Trello. Something else that's exciting, I've got a conversation later with a representative from Manning Publications. They do a lot of books on technology. In fact, if you've done the Deep Learning Nanodegree, you might have read Andrew Trask's Rocking Deep Learning, which is through Manning. They reached out to me last week, you might have seen in last week's episode, to review one of their book proposals. And then as a follow-up from that, when I replied to that, they said, can they contact me and have a conversation about what I've learned in the AI Nanodegree/Deep Learning Nanodegree? And of course, guys, I still treat myself as a novice, but I'm excited to talk to this guy and see what he wants. After all, I do really enjoy writing. I do really enjoy any type of communication. So if an opportunity appears for that, I'd be all over that. But otherwise, if not, that's all good. I'm happy to keep going with what we're doing. I'm happy of what we're doing here, what we're building here. And after the conversation, actually, that's in about 2 hours' time, I'll check in and tell you how it goes. So I might get a little bit I might have to go for a walk beforehand. That's a tip, by the way. If you've got to talk or you have some sort of presentation or some sort of interview, one of my two tips is go for a walk beforehand, 20 minutes, or do a warm-up conversation with a friend on the phone, because that way it'll get the blood— well, get the conversation juices flowing so that you're not so much nervous. I just finished up a conversation with Greg. So, before in the last clip, I said I was having a conversation with someone from Manning Publications, and it was very nice to meet you, Greg. So, essentially what happened is they'd stumbled upon my YouTube channel and were reaching out to find out what I've been learning in my artificial intelligence Nanodegree and the deep learning Nanodegree and just a bit more about my journey and where I'm up to and whatnot. And they were interested and wondering whether I'd be looking to make or go in partnership with them in the future and make some videos for their platform. They're an online learning company, so they've got a video platform that's been up for about 6 months or so, but I think it's only invite access at this stage. And they, of course, sell or have a massive listing of some really great books. And I've got a few that I've read and definitely learned some great things. From. I'll let you all know if something happens in the future, but for now we're going to keep doing what we're doing. If you want to see any videos in particular, don't forget, leave a comment below. But otherwise, tomorrow I'm going to keep continuing on with the artificial intelligence allegory. Today was all about planning of Project 4. Planning is done and now it's just diving in and doing it. We'll get into some coding tomorrow and we'll see where we're at. I made some progress, guys. So the 4 dots here means passed. So we ran 4 tests and we got an okay back from the, from the unit test module. Part 1 of Project 4 is, is complete, or complete that thus far. I might have to go back over it in the future and figure out if I can improve it in any way, shape, or form. Otherwise, we're up to part 2, and part 2 is model selection. So we're going to use one of these 3 methods here: log likelihood using cross-validation folds, or CV, Bayesian Information Criterion, BIC, or discriminative information criterion, DIC. And I'll be upfront with you, I'm not entirely sure of what any of those mean, so that's going to be tomorrow's task of lift myself up a little to-do list of where to start, of what to do next, just so tomorrow I can get straight into it. Like, that's what I like to do. If I'm finishing up a section or finishing up work for the day, I like to leave myself like a little continuation on for the next day. So when I go to sit down, often that's the— or start work, that's the hardest point, right? It's just starting. So if you leave yourself a little hint, a little trail that you can follow and get back into it, be much better off for it. Otherwise, I started looking over one of them the BIC. It's got this formula here, supposed to implement that in Python. That's task 1 for tomorrow. I made some more progress on project 4 today, guys. About halfway through part 2. I like to say halfway even though we've got a few things to do. So I've got one, one of the main algorithms out of the way, and usually based off past history, once I get one of them out of the way, the, the knowledge and the momentum carries me over and allows me to get the other ones done faster. But I'm done for coding for today. I'm gonna go to the gym and celebrate it. I've done about, I think, 6 or 7 Pomodoros on just pure coding today. I started off with HackerRank. If you haven't seen HackerRank before, it's really great to practice your coding in general, algorithms, different languages, data structures, all that kind of stuff. So I'm practicing Python on there. I'm trying to do at least 25 minutes every day just dedicated to just implementing or just attempting the Python challenges. And I've got it on the Trello board as well, so you'll be able to see that in the link in the description. Otherwise, next week's video is going to be me completing Project 4. Well, it should be as long as things go smoothly. And otherwise, yeah, they will. Let's just, let's just say that. Next video is me completing Project 4, so that'll be the— that'll wrap up Term 1 of the Artificial Intelligence Nanodegree, which is exciting. That means I get to have some sort of little— I think there's a little gap in between Term 1 and Term 2. Then we'll work out what we're doing in the interim, but the videos will still keep going. I think I'm going to start some sort of new course, uh, in the interim to, to optimize that time, optimize that free time, because I don't want to lose the momentum I'm gathering up. As we're moving through this. I've got some thank yous to give out to the people who have reached out to me via email. If you ever want to do, feel free to. My email is daniel@mrdbourke. I respond to all emails, and if I don't, just give me some time, or otherwise just hit me up again and prompt me to do. I try to answer questions about anything to do with fitness, nutrition, studying AI. Remember, I'm a complete newbie to programming and artificial intelligence still. Bear with me, I'll do my best, but I can't guarantee that the answer is is 100% correct. If I don't have the right answer, I'll try and direct you to someone who I think will be able to better answer the question for me— or for you, sorry. Shoutouts of the week, we got Jack, Jose, Ben, Andrew, Anoush, and Verdu. Thank you so much for reaching out, guys. Thank you for your kind words. Hope I did my best to answer your questions. If you have any more, hit me up always, and that goes for anyone else. Or leave a comment below if you want to see a future video. Email me, hit me up on Twitter, or send a carrier pigeon my way. And last but not least, A big shout out to Karen, my first Patreon. As you may or may not know, guys, I've got a Patreon page. I'll put that— the link is in the description. As always, I'll keep making videos, but if they're not bringing you value, please send me an email, give me some advice on how I can make them better. But otherwise, subscribe, leave the video a like, and we'll catch you next week.