1 00:00:02,350 --> 00:00:08,930 This is what is known in software engineering as test driven development or TDD in test driven development. 2 00:00:08,950 --> 00:00:13,690 We start by running tests because number one, your code should always have tests. 3 00:00:14,140 --> 00:00:19,960 And number two, running the test first forces you to think about how you expect your API to work. 4 00:00:20,320 --> 00:00:25,990 And it allows you to make sensible design decisions without being distracted by implementation details. 5 00:00:26,770 --> 00:00:31,330 So you see how this applies to every machine learning algorithm, not just linear regression. 6 00:00:35,610 --> 00:00:40,980 And a lot of my older coding videos, I unfortunately decided to just write the code straight through. 7 00:00:41,550 --> 00:00:44,220 This helps some people, but is not ideal for others. 8 00:00:44,760 --> 00:00:47,850 In addition, the coding videos will appear without warning. 9 00:00:48,600 --> 00:00:51,120 Generally, the lectures appear in the same pattern. 10 00:00:51,600 --> 00:00:54,450 First, we do theory and then we do the corresponding code. 11 00:00:54,870 --> 00:00:56,670 It pretty much always works this way. 12 00:00:57,840 --> 00:01:02,910 So after the theory lecture, you can be pretty sure we are going to implement that theory in code soon, 13 00:01:03,270 --> 00:01:09,510 unless it's so simple that it can be an exercise now to avoid treating everyone like they're in kindergarten. 14 00:01:09,750 --> 00:01:12,600 I don't have videos telling you that it's time to code. 15 00:01:13,200 --> 00:01:15,660 You're all proactive and independent adults. 16 00:01:15,900 --> 00:01:21,180 So when you learn some theory, you already know that it's time to try to put that into code by yourself. 17 00:01:25,580 --> 00:01:30,350 Sometimes you might get stuck and have no idea what's a writing code or what the data set is. 18 00:01:30,920 --> 00:01:35,660 In that case, it's fine to watch the next coding lecture, but don't watch it all the way through. 19 00:01:36,290 --> 00:01:41,450 Watch just enough so that you know what the general idea is and what we're trying to accomplish and 20 00:01:41,450 --> 00:01:43,220 then try to finish the rest yourself. 21 00:01:48,000 --> 00:01:51,240 Now, let's say you've watched just enough so that you know what we're trying to do. 22 00:01:51,720 --> 00:01:55,990 You know the theory, but you still can't complete the code at this point. 23 00:01:56,010 --> 00:02:00,450 You may want to revisit the lecture theory to make sure you really understood each part. 24 00:02:01,260 --> 00:02:05,310 Failing that, you may want to try and get clarification on the discussion board. 25 00:02:05,910 --> 00:02:09,210 Failing that, you may want to look at external resources as well. 26 00:02:09,990 --> 00:02:14,880 Sometimes you might be missing some important computer science background that I've assumed everyone 27 00:02:14,880 --> 00:02:15,690 already knows. 28 00:02:16,260 --> 00:02:19,410 That's perfectly fine because everyone has a different background. 29 00:02:20,130 --> 00:02:24,480 It's highly encouraged to do your own research, and this is also perfectly normal. 30 00:02:24,930 --> 00:02:25,980 It's not a bad thing. 31 00:02:25,980 --> 00:02:29,010 It's a good thing, as can a university student. 32 00:02:29,010 --> 00:02:32,160 If this is what they do and they'll tell you that they do it all the time. 33 00:02:32,940 --> 00:02:36,120 This builds a level of mental toughness and initiative taking. 34 00:02:36,360 --> 00:02:38,580 That's important if you want to learn this stuff. 35 00:02:39,210 --> 00:02:42,960 There is a lot to learn, and I myself am still learning new things every day. 36 00:02:46,360 --> 00:02:51,400 So let's say you've rewatched the theory lecture, you've looked at some external resources on the theory, 37 00:02:51,670 --> 00:02:54,370 and you still don't understand how you would put it into code. 38 00:02:55,210 --> 00:03:00,520 Well, at this point, there are some great exercises you can do watch the coding lecture or look at 39 00:03:00,520 --> 00:03:01,720 the full code on GitHub. 40 00:03:02,380 --> 00:03:04,840 Go through it and make sure you understand each line. 41 00:03:05,860 --> 00:03:10,210 The next step is without looking at the code again, try to implement it yourself. 42 00:03:10,870 --> 00:03:12,880 This by itself is a great learning tool. 43 00:03:13,510 --> 00:03:18,670 Many of the algorithms I've learned, I didn't understand them the first time I learned them by looking 44 00:03:18,670 --> 00:03:21,340 at someone else's implementation and dissecting it. 45 00:03:23,240 --> 00:03:28,220 I looked at each part and I asked myself which parts that I get right and which parts did I get wrong? 46 00:03:29,150 --> 00:03:32,180 This helps you understand the mistakes you were making previously. 47 00:03:35,760 --> 00:03:39,120 The cool thing about implementation is it works backwards, too. 48 00:03:39,720 --> 00:03:45,270 We go from theory to code, but thinking in the reverse direction, the code actually helps us understand 49 00:03:45,270 --> 00:03:46,440 the theory better also. 50 00:03:47,250 --> 00:03:51,360 I always say that if you can't implement the code, then you don't understand the theory. 51 00:03:52,140 --> 00:03:57,900 So if you can learn the theory, 50 percent of the way, try to implement the code, look at an existing 52 00:03:57,900 --> 00:04:00,240 implementation and then finally get it working. 53 00:04:00,510 --> 00:04:04,830 Then you can be guaranteed that you now know more than just 50 percent of the theory. 54 00:04:05,460 --> 00:04:10,320 The act of coding helps you understand more of the theory than if you had just learned the theory by 55 00:04:10,320 --> 00:04:10,800 itself. 56 00:04:14,130 --> 00:04:19,440 So implementation helps you understand the theory, but it also helps you understand the intuition as 57 00:04:19,440 --> 00:04:19,800 well. 58 00:04:20,579 --> 00:04:22,230 Intuition is a tricky thing. 59 00:04:22,320 --> 00:04:27,900 In fact, I would go so far as to say it's a trap in many, or maybe even most cases. 60 00:04:27,960 --> 00:04:29,970 Intuition is built on experience. 61 00:04:30,540 --> 00:04:35,550 That means deriving the theory from first principles or, in other words, deriving the theory from 62 00:04:35,550 --> 00:04:38,970 facts you already know to be true and from implementation. 63 00:04:39,900 --> 00:04:43,740 Students often want to hear intuition from people with lots of experience. 64 00:04:44,070 --> 00:04:47,340 But the problem is that this intuition is built on hindsight. 65 00:04:51,200 --> 00:04:55,640 Intuition is often both the first thing and the last thing you want to focus on. 66 00:04:56,570 --> 00:05:01,730 You want to have a little bit of intuition at the beginning, so you have some general idea of what 67 00:05:01,730 --> 00:05:02,870 you want to accomplish. 68 00:05:03,590 --> 00:05:08,930 Then after you learn the theory and implementation, you can reflect on the details and summarize them 69 00:05:08,930 --> 00:05:09,680 at a high level. 70 00:05:10,310 --> 00:05:13,540 This is real intuition built on hindsight and experience. 71 00:05:18,350 --> 00:05:23,690 The big mistake beginners make is that they read some intuition written by someone else who actually 72 00:05:23,690 --> 00:05:26,090 has experience when they themselves do not. 73 00:05:27,080 --> 00:05:30,110 From an outsider's perspective, it's hard to tell the difference. 74 00:05:30,680 --> 00:05:31,430 Why is that? 75 00:05:32,570 --> 00:05:37,730 Well, if the beginner can regurgitate the same high level ideas as the expert, then it would sound 76 00:05:37,730 --> 00:05:40,280 like they are also an expert when in fact they are not. 77 00:05:41,320 --> 00:05:44,470 There are, unfortunately, too many courses out there that do this. 78 00:05:45,160 --> 00:05:49,930 They want you to feel good that you learn something and they want to avoid you having to do anything 79 00:05:49,930 --> 00:05:50,650 too difficult. 80 00:05:51,670 --> 00:05:56,530 And because you can repeat the same high level intuitions, it actually sounds like you're an expert. 81 00:05:57,100 --> 00:05:58,720 So this is why I call it a trap. 82 00:06:03,340 --> 00:06:06,700 Knowing intuition not only makes you feel like, you know, something. 83 00:06:07,060 --> 00:06:09,610 It also makes you really confident about this fact. 84 00:06:10,300 --> 00:06:16,060 And then because your ego wants to maintain this confidence, you'll start making crazy statements like 85 00:06:16,060 --> 00:06:21,160 the theory and implementation are not important because you already understand the intuition. 86 00:06:21,760 --> 00:06:24,100 But as you know, this is completely backwards. 87 00:06:28,150 --> 00:06:34,150 Let's summarize this lecture, since that was a lot of information to take in, remember, this lecture 88 00:06:34,150 --> 00:06:36,820 is all about how and why you should code by yourself. 89 00:06:37,450 --> 00:06:43,270 First, we talked about why this is because it ensures you are thinking about the details and filling 90 00:06:43,270 --> 00:06:46,000 in the blanks using your own pattern recognition skills. 91 00:06:46,690 --> 00:06:51,130 It ensures you don't just assume you know what's going on, but that you actually know what's going 92 00:06:51,130 --> 00:06:51,460 on. 93 00:06:52,210 --> 00:06:56,590 There's a big difference, especially if you're the type of person that likes to assume, you know, 94 00:06:56,590 --> 00:06:57,160 everything. 95 00:06:58,090 --> 00:07:04,090 Next, we talked about how we talked about how all data is the same, and it doesn't really matter what 96 00:07:04,090 --> 00:07:06,100 domain or industry your data is from. 97 00:07:06,370 --> 00:07:08,110 The same algorithms all apply. 98 00:07:08,950 --> 00:07:14,140 Next, we looked at the interfaces to the algorithms, which are also conveniently all the same. 99 00:07:14,800 --> 00:07:20,770 So we have a kind of mix and match type of scenario where we can input any kind of data into any kind 100 00:07:20,770 --> 00:07:21,280 of model. 101 00:07:22,160 --> 00:07:28,010 For example, you can use linear regression with financial data or you can use a neural network on ecological 102 00:07:28,010 --> 00:07:28,370 data. 103 00:07:28,910 --> 00:07:32,420 All data is the same in all, models have the same interface. 104 00:07:36,310 --> 00:07:39,910 Next, we talked about a strategy for how to actually apply this. 105 00:07:40,660 --> 00:07:44,650 You've probably already seen that we always alternate between theory and code. 106 00:07:45,280 --> 00:07:48,580 First, we discuss the theory and then we implement that in code. 107 00:07:49,480 --> 00:07:54,340 Sometimes you might have to go back and revisit the theory, or sometimes you might have to look things 108 00:07:54,340 --> 00:07:55,180 up on the internet. 109 00:07:55,630 --> 00:07:56,830 That's perfectly normal. 110 00:07:57,870 --> 00:08:00,960 This is pretty much the day in the life of a university student. 111 00:08:01,680 --> 00:08:06,270 We're trying to build the same level of mental toughness and initiative taking in this course. 112 00:08:06,990 --> 00:08:11,760 So the strategy is stop after you learn the theory and try to code yourself. 113 00:08:12,710 --> 00:08:17,600 Sometimes you need a little more detail, so you might want to watch a little bit of the coding lectures 114 00:08:17,600 --> 00:08:20,930 to see what's going on and then try to implement the rest yourself. 115 00:08:21,680 --> 00:08:26,570 If you can't get it by yourself, then maybe try watching the whole thing or looking at the expected 116 00:08:26,570 --> 00:08:29,330 result and then trying to implement it from memory. 117 00:08:30,140 --> 00:08:34,850 This by itself is a great learning tool because it allows you to ask yourself, What did I get right 118 00:08:34,850 --> 00:08:35,990 and what did I get wrong? 119 00:08:36,260 --> 00:08:38,419 So you can understand the mistakes you were making. 120 00:08:42,820 --> 00:08:48,520 Finally, it's important to understand the difference between intuition, theory and implementation 121 00:08:48,670 --> 00:08:50,320 and how they reinforce each other. 122 00:08:51,070 --> 00:08:56,740 It's important to not fall into the trap of only understanding intuition written by someone else and 123 00:08:56,740 --> 00:09:00,160 then assuming that this equates to actually understanding everything. 124 00:09:01,030 --> 00:09:06,250 In many cases, it's not about waiting for someone to tell you some nice phrases that trick you into 125 00:09:06,250 --> 00:09:11,710 thinking you understand the intuition, but rather you should seek intuition by working through the 126 00:09:11,710 --> 00:09:13,300 theory and the implementation. 127 00:09:13,900 --> 00:09:18,010 And lastly, don't forget when it's time to code, you must code.