MACHINE LEARNING IN 5 MINUTES

Hello world, what's going on? Today we're going to learn about machine learning, or more specifically, just a brief overview of what is machine learning in about 5 or so minutes. So we're going to cover 4 main things. What is machine learning, of course, and a little overview. Number 2, types of machine learning. What are the main types of machine learning out there? There's plenty, but we'll cover the 3 main ones. Number 3, where is machine learning being used right now? And number 4, where can you go to learn machine learning if you wanted to learn about this stuff? So number 1, what is machine learning? So in a general overview, machine learning is the idea that you can have a computer do something for you without it being explicitly programmed to do that. So for the majority of computing history, you've had to hardcode in instructions for a computer to do something. But now with this new era of machine learning and all the data we've been collecting for years and years, now machines can learn from that data and make decisions for themselves, effectively becoming artificial intelligence. So like you or I can walk into a room and we don't have to be explicitly told what to do in the room, we just sort of generally get it. That's the end goal of machine learning, is to have machines be able to recognize situations and be able to make decisions for themselves. So number 2, what types of machine learning are there? Well, I'm going to go over 3 main types, but just beware, there's plenty more. So supervised learning, unsupervised learning, and reinforcement learning. What's an example of each? So supervised learning, say for example, you had a whole bunch of data on house selling prices, and you knew the number of rooms, square feet, and suburb that the houses were in, and their selling price. You could train a machine to take in all that data and then produce out the future selling prices of houses if you knew their number of rooms, square feet, and suburb that they're in. You can deduce that selling price using machine learning. So unsupervised learning, what's an example of that? Well, say you had that big dataset as well of all the house prices, room numbers, square feet, and whatnot, but you wanted to find some other patterns. Well, you could use a machine to go through that data and find out patterns that you wouldn't really recognize yourself unless you spent years and years working on it. See, machines are a lot faster at going through big amounts of data than humans are. Number 3 is reinforcement learning. So, reinforcement learning can be similar to training a dog, right? If the dog does good behavior, you reward the dog for that good behavior. If it doesn't do good behavior, then you don't reward the dog. So take for example the example of a self-driving car. If the car crashes, then you change the algorithm that caused the crash. But if the car drives through the lanes and sticks to the lanes, well, you strengthen those algorithms. So there's 3 main types of machine learning. 3. So where is machine learning being used? Everywhere. The device I'm recording on right now more than likely uses machine learning to recognize my face and focuses in on that area. The device you're watching this on more than likely uses machine learning to improve its battery life. You go to Google, every single product at Google is using machine learning. Gmail, YouTube, everything. Facebook, if you use Facebook, you scroll through your newsfeed, if you like someone's post, the machine will learn, the machine learning algorithm that is behind Facebook's newsfeed will learn that you like their post. and so you'll see more from that person. So where's machine learning being used? Everywhere, and it's only going to get more popular as technology improves. Number 4. So where can you learn machine learning? Where can you get a butt on this hype train of— well, it's not really a hype train, it's going to be here for a long time. So where can you find out more about this stuff? So for starters, I'd probably recommend Coursera. Machine learning, the course on there is probably the best one available for free right now. It's by Stanford University and Andrew Ng. I can't really pronounce his last name properly, but amazing course. I'm doing it right now. I'm 4 weeks into it. Udacity is another awesome resource for machine learning. You probably want to learn some Python code. So anywhere like Treehouse or Code School, learning about Python and statistics and algebra. So I'd go to Khan Academy for that. Also, Siraj Raval has an incredible YouTube channel. Check him out. He's also far more entertaining than I am. And finally, I'd say if you want to read more, go to the machine learning subreddit. I find a lot of cool articles in there. They're quite complex, but it'll give you some idea of what you're learning. So that's it. What is machine learning? A brief overview. Of course, there's so much more to learn. I'll put all the links I talked about in the description so you can learn more. Otherwise, stay Stay tuned, I'll have more videos in the future on AI, machine learning, fitness, nutrition, etc. And just generally whatever I'm up to. But thank you so much for watching and have an awesome day!