1 00:00:08,170 --> 00:00:12,430 Hey guys what's going on this is Caleb with slopes and in this video we're going to be talking about 2 00:00:12,430 --> 00:00:13,470 machine learning. 3 00:00:13,470 --> 00:00:19,450 We're going to just kind of do a brief umbrella discussion about machine learning at large. 4 00:00:19,450 --> 00:00:26,020 Before we dive into the code and actually start using Core M-L in our apps First we need to understand 5 00:00:26,200 --> 00:00:28,040 what is machine learning. 6 00:00:28,060 --> 00:00:33,420 We also are going to talk a bit about how it works and why should we be using machine learning. 7 00:00:33,440 --> 00:00:34,240 OK what's the point. 8 00:00:34,250 --> 00:00:35,520 Why why should we do it. 9 00:00:35,710 --> 00:00:42,790 But first let's go and just talk about what is machine learning machine learning is basically a way 10 00:00:42,820 --> 00:00:51,130 that our programs on our applications can learn from various data that's been passed into our application 11 00:00:51,550 --> 00:00:54,260 and then make predictions based on that information. 12 00:00:54,340 --> 00:01:01,120 In the example of the app that you'll build in this course we use core M-L we use machine learning to 13 00:01:01,120 --> 00:01:04,740 classify information that comes in through a photo. 14 00:01:04,900 --> 00:01:09,060 Then our our model which we'll talk about the trained model later. 15 00:01:09,250 --> 00:01:12,940 Basically the model is going to take that input data as a photo. 16 00:01:12,940 --> 00:01:19,150 It will analyze patterns it can see in the photo and then it can use that information to make a prediction 17 00:01:19,570 --> 00:01:21,300 based on what it already knows. 18 00:01:21,460 --> 00:01:22,860 So that's pretty cool. 19 00:01:23,170 --> 00:01:26,630 A definition I really like of machine learning is right here. 20 00:01:26,690 --> 00:01:27,170 OK. 21 00:01:27,270 --> 00:01:30,540 And it's pretty vague but I'm going to go into detail about it later. 22 00:01:30,730 --> 00:01:37,830 Machine learning gives computers the ability to learn without being explicitly programmed. 23 00:01:37,840 --> 00:01:44,580 So I did not have to write the code to tell my app how to identify a remote or a window or an umbrella. 24 00:01:44,590 --> 00:01:46,590 I didn't have to tell it to do that. 25 00:01:46,780 --> 00:01:48,710 I used a pre-trained model. 26 00:01:48,730 --> 00:01:50,440 Think of that like a brain. 27 00:01:50,530 --> 00:01:56,810 When you are a baby you have a brain obviously but it's still developing it's still learning and growing. 28 00:01:56,830 --> 00:02:01,650 So you need to have experiences you need to be exposed to different kinds of information. 29 00:02:01,660 --> 00:02:02,920 You need someone to teach you right. 30 00:02:02,920 --> 00:02:05,810 Maybe your parents teach you oh this is hot right. 31 00:02:05,830 --> 00:02:06,900 Don't touch the stove. 32 00:02:07,030 --> 00:02:12,040 They tell you that it's hot and you probably know that but you don't understand and don't experience 33 00:02:12,040 --> 00:02:13,310 it until you touch the cell. 34 00:02:13,330 --> 00:02:14,370 Wow that's hot. 35 00:02:14,440 --> 00:02:16,390 And you make that connection in your brain. 36 00:02:16,480 --> 00:02:18,570 Pre-trained model is kind of like that. 37 00:02:18,710 --> 00:02:19,130 OK. 38 00:02:19,150 --> 00:02:25,540 Data scientists are kind of like parents in a way because they train these these machine learning models 39 00:02:25,840 --> 00:02:29,250 to understand information and to classify it. 40 00:02:29,410 --> 00:02:32,670 So it's exposed to a lot of different kinds of information. 41 00:02:32,800 --> 00:02:39,290 Maybe the data scientists expose their model to hundreds of different pictures of TV remotes. 42 00:02:39,400 --> 00:02:46,330 So that the model can identify OK when I see a shape that's about this size that has a bunch of square 43 00:02:46,330 --> 00:02:51,520 buttons in a certain grid and order with different colors it knows to classify that and observe those 44 00:02:51,520 --> 00:02:56,500 patterns and then based on that it can make a prediction with a certain level of confidence. 45 00:02:56,590 --> 00:02:57,910 That's really cool. 46 00:02:57,940 --> 00:03:05,620 Now basically the computer has been trained it's learned what makes a remote What makes an umbrella. 47 00:03:05,710 --> 00:03:11,170 And based on its experience and what it's learned it's able to make a prediction very very cool. 48 00:03:11,170 --> 00:03:16,840 This technology of computers being able to learn without being explicitly programmed is pretty incredible 49 00:03:16,840 --> 00:03:22,450 in my opinion and the implications of how machine learning could change the world and are changing the 50 00:03:22,450 --> 00:03:26,630 world go beyond just simple apps that classify photos. 51 00:03:26,680 --> 00:03:32,350 Doctors are using machine learning to identify and diagnose certain diseases. 52 00:03:32,360 --> 00:03:33,110 OK. 53 00:03:33,490 --> 00:03:38,770 It can even be used for like relationship stuff it can be used to identify patterns and habits maybe 54 00:03:38,770 --> 00:03:41,350 for like relationship or matchmaking. 55 00:03:41,350 --> 00:03:48,760 It could be used to analyze very expensive and costly manufacturing procedures like will this engine 56 00:03:48,820 --> 00:03:54,420 work or will a certain part fail it can make predictions based on how it has been trained. 57 00:03:54,520 --> 00:03:59,390 There's lots of very cool ways in the real world that machine learning could be used. 58 00:03:59,620 --> 00:04:00,370 OK. 59 00:04:00,370 --> 00:04:05,780 There are a couple different ways that machine learning works and now we're into the how does it work. 60 00:04:05,800 --> 00:04:06,960 Part of this video. 61 00:04:07,210 --> 00:04:13,720 So there is something called supervised machine learning where basically when the model is being trained 62 00:04:14,170 --> 00:04:20,560 the data scientists or whomever is programming it they might show it a picture of a dog and say OK this 63 00:04:20,560 --> 00:04:21,260 is a dog. 64 00:04:21,370 --> 00:04:27,190 And the model knows OK that's a dog great then it might show a photo of a cat and say OK this is not 65 00:04:27,190 --> 00:04:27,830 a dog. 66 00:04:28,000 --> 00:04:28,660 All right. 67 00:04:28,660 --> 00:04:30,190 And so then the model can tell. 68 00:04:30,190 --> 00:04:31,270 OK this is a dog. 69 00:04:31,270 --> 00:04:32,370 This is not a dog. 70 00:04:32,560 --> 00:04:39,270 Now the issue though is that supervised M-L requires a lot of human input to verify that the model is 71 00:04:39,340 --> 00:04:40,830 is doing it correctly. 72 00:04:41,030 --> 00:04:41,640 OK. 73 00:04:41,910 --> 00:04:46,630 OK so since the model is being supervised the human is saying yep that's a dog. 74 00:04:46,660 --> 00:04:47,080 OK. 75 00:04:47,140 --> 00:04:48,520 This is not a dog. 76 00:04:48,550 --> 00:04:49,520 This is a cat. 77 00:04:49,540 --> 00:04:50,350 This is a dog. 78 00:04:50,350 --> 00:04:51,670 This is the dog. 79 00:04:51,760 --> 00:04:59,260 The model can learn and identify patterns and then based on that in the future it can classify new data 80 00:04:59,260 --> 00:05:04,690 it receives and make a prediction based on what it has already seen of cats and dogs that could analyze 81 00:05:04,690 --> 00:05:11,170 the photo and realize OK so some dogs are very big Some dogs are very small but most cats are about 82 00:05:11,170 --> 00:05:12,240 the same size. 83 00:05:12,250 --> 00:05:12,860 OK. 84 00:05:13,050 --> 00:05:17,410 A cat has a smaller nose than most dogs and a dog has a bigger nose than most cats. 85 00:05:17,410 --> 00:05:21,530 So based on those patterns it can see in the photo it can make predictions. 86 00:05:21,550 --> 00:05:22,530 Very cool. 87 00:05:23,650 --> 00:05:30,010 Be unsupervised machine learning there's also unsupervised machine learning and based on some inherent 88 00:05:30,010 --> 00:05:36,670 characteristics of those images passed in the model can start to identify patterns and make its own 89 00:05:36,670 --> 00:05:39,080 predictions in its own classifications. 90 00:05:39,070 --> 00:05:40,930 Kate that's very very cool. 91 00:05:41,140 --> 00:05:46,900 And there are of course a couple of different ways that data is structured in machine learning models 92 00:05:46,930 --> 00:05:52,810 but most of the time it will be structured in either a decision tree where basically one decision yes 93 00:05:52,810 --> 00:05:57,220 or no lead you to another part of the tree yes or no lead you to another part of the tree and so on 94 00:05:57,220 --> 00:06:02,390 and so forth until it can make a prediction with a fairly high level of confidence. 95 00:06:02,410 --> 00:06:09,660 Beyond that there are artificial neural networks which basically connect data that is similar or related 96 00:06:09,820 --> 00:06:16,030 case similar to how our brain works with different synopses and different nodes in our brains for creating 97 00:06:16,030 --> 00:06:21,040 thoughts and memories and artificial neural network works just like that where it makes connections 98 00:06:21,040 --> 00:06:26,020 and it makes observations based on patterns and related data that it already knows about. 99 00:06:26,080 --> 00:06:26,770 Very very cool. 100 00:06:26,770 --> 00:06:31,750 Those are just two common ways that data is structured for a machine learning model. 101 00:06:32,020 --> 00:06:36,100 And now we should talk about why use machine learning in our apps what's the point. 102 00:06:36,160 --> 00:06:37,260 Why should I use this. 103 00:06:37,420 --> 00:06:42,730 And I'm going to tell you the number one reason is that it makes it easy for you to personalize your 104 00:06:42,730 --> 00:06:44,410 app for your user. 105 00:06:44,650 --> 00:06:50,860 If you integrate machine learning you can add features to your app that pay attention to what your user 106 00:06:50,860 --> 00:06:54,350 does and make suggestions and predictions based on that. 107 00:06:54,610 --> 00:07:01,630 If you've ever shopped on Amazon.com you know when things pop up that say recommended for you or based 108 00:07:01,630 --> 00:07:04,510 on this we think you might like this other product. 109 00:07:04,510 --> 00:07:09,580 It uses a bunch of data from people buying things all over the Internet and then based on what other 110 00:07:09,580 --> 00:07:14,340 people have purchased it can make a recommendation to you because of your shopping habits. 111 00:07:14,440 --> 00:07:19,300 It finds people like you that buy things like you and recommends things that you might like that's machine 112 00:07:19,300 --> 00:07:19,980 learning. 113 00:07:20,000 --> 00:07:26,290 OK there's lots of great applications for this to make your apps tailored specifically to your user. 114 00:07:26,290 --> 00:07:30,160 The coolest thing is that machine learning is easy and available to use. 115 00:07:30,160 --> 00:07:32,860 Now for IO West that's what we're going to learn in this course. 116 00:07:32,860 --> 00:07:34,390 And it's so cool. 117 00:07:34,390 --> 00:07:39,760 So let's go ahead let's stop talking about machine learning at large and let's dive into core Armel 118 00:07:39,790 --> 00:07:40,930 for iOS. 119 00:07:40,930 --> 00:07:42,640 We'll see in the next video.