1 00:00:00,510 --> 00:00:05,970 Highlight continue with our cause and this video, we will do the multiple linear regression model. 2 00:00:06,360 --> 00:00:11,610 So multiple linear regression is a straightforward generalisation of single predictive models. 3 00:00:11,910 --> 00:00:18,150 In a multiple linear regression model, the dependent variable is related to two or more independent 4 00:00:18,150 --> 00:00:22,030 variables to perform on multiple linear regression analysis. 5 00:00:22,530 --> 00:00:31,110 So the cycle library where we use so from the scale not linear scale model, feel the linear regression 6 00:00:31,110 --> 00:00:37,080 clasper, formerly an ordinary at least square linear regression. 7 00:00:37,560 --> 00:00:39,750 So as usual, we will load the library. 8 00:00:41,790 --> 00:00:43,920 By typing in here. 9 00:00:44,930 --> 00:00:48,800 And from Escalon Dot. 10 00:00:50,300 --> 00:00:51,270 The new model. 11 00:00:52,530 --> 00:00:58,170 Import linear regression and then. 12 00:00:59,400 --> 00:01:02,460 Now we lose you to our model. 13 00:01:04,100 --> 00:01:08,450 So is it right now moral equal? 14 00:01:10,550 --> 00:01:11,030 The. 15 00:01:11,920 --> 00:01:14,020 Near recession. 16 00:01:16,580 --> 00:01:21,140 And then to fit the millennium model where you would fit function. 17 00:01:22,110 --> 00:01:24,090 So, Al. 18 00:01:25,030 --> 00:01:25,780 Modahl. 19 00:01:27,600 --> 00:01:28,350 Don't fit. 20 00:01:30,550 --> 00:01:35,200 We actually underscore a train come, why underscore train? 21 00:01:38,360 --> 00:01:44,960 And then it this guy in the training phase with you, the data instructed fodder for this fire and at 22 00:01:44,960 --> 00:01:47,470 this point we can build the model to make that prediction. 23 00:01:48,440 --> 00:01:50,890 So it's very easy to predict my thoughts. 24 00:01:51,110 --> 00:01:56,550 So in here, I ask you, why are those Brad Alam? 25 00:01:58,840 --> 00:02:00,490 Aguer, al. 26 00:02:01,600 --> 00:02:02,290 Model. 27 00:02:04,070 --> 00:02:05,330 Don't predict. 28 00:02:07,130 --> 00:02:08,140 And that's test. 29 00:02:12,540 --> 00:02:17,520 And they told us how so I hope I don't make any mistake. 30 00:02:19,720 --> 00:02:25,770 So usually, I scatterplot is you to determine whether or not there is a relationship between data, 31 00:02:25,990 --> 00:02:33,130 however, as catalog can be used to analyze the performance of a linear model by reporting the actual 32 00:02:33,130 --> 00:02:35,840 and predictive value on the two assets. 33 00:02:35,860 --> 00:02:40,600 Is it possible to check how the data is arranged to help with the analysis? 34 00:02:40,610 --> 00:02:43,630 Is it possible to trace the bisect the Cuarón? 35 00:02:44,050 --> 00:02:53,290 That is the line of equation Wilko X theoretically on observations to rest on this slide, but we can 36 00:02:53,440 --> 00:03:01,750 be satisfied that the data is closer to this line about how the data must be below the line and the 37 00:03:01,750 --> 00:03:03,630 other half must be about the data. 38 00:03:03,970 --> 00:03:10,780 So the boys that move away significantly from the life of these libraries and a possible outliners so 39 00:03:10,780 --> 00:03:19,510 that Blossom's scatterplot using the library and to do so will pay out a lot. 40 00:03:20,290 --> 00:03:22,240 Figure one. 41 00:03:23,710 --> 00:03:24,730 Be out, they don't. 42 00:03:26,500 --> 00:03:27,190 Subplot. 43 00:03:30,050 --> 00:03:31,520 One, two, one. 44 00:03:33,040 --> 00:03:35,340 Partido Scotter. 45 00:03:39,240 --> 00:03:40,410 Why underscore? 46 00:03:41,410 --> 00:03:50,140 That's why underscore bread and then potato. 47 00:03:53,710 --> 00:03:56,080 Label with a. 48 00:03:59,360 --> 00:04:00,170 Actual. 49 00:04:02,350 --> 00:04:03,160 Values. 50 00:04:04,750 --> 00:04:06,660 And then out the door. 51 00:04:07,180 --> 00:04:07,780 Why? 52 00:04:08,850 --> 00:04:09,510 Label. 53 00:04:11,800 --> 00:04:12,580 With a. 54 00:04:14,660 --> 00:04:14,920 That. 55 00:04:16,880 --> 00:04:22,130 Values, and the last one is a dog title. 56 00:04:23,860 --> 00:04:24,510 Is a. 57 00:04:25,890 --> 00:04:30,120 New role network model. 58 00:04:33,350 --> 00:04:37,820 And yet you hear the dog subplot. 59 00:04:39,460 --> 00:04:47,280 One to one, I actually went to two and pay out a door, Scotter. 60 00:04:48,920 --> 00:04:52,040 They lie under is called test. 61 00:04:55,110 --> 00:04:57,340 Why underscore, right? 62 00:04:59,590 --> 00:05:01,590 Be out the door, I. 63 00:05:03,140 --> 00:05:04,390 Label is. 64 00:05:07,240 --> 00:05:08,060 After. 65 00:05:10,790 --> 00:05:13,340 Well, you played on. 66 00:05:15,610 --> 00:05:17,590 Why label is. 67 00:05:20,280 --> 00:05:20,970 Predict that. 68 00:05:22,420 --> 00:05:23,290 Values. 69 00:05:27,790 --> 00:05:28,840 They are they don't. 70 00:05:30,630 --> 00:05:34,920 Title with a české. 71 00:05:36,480 --> 00:05:36,990 Lord. 72 00:05:39,550 --> 00:05:40,450 Lydia. 73 00:05:45,580 --> 00:05:48,340 Gray chant model. 74 00:05:49,250 --> 00:05:50,900 And then data show. 75 00:05:52,490 --> 00:05:53,480 Very simple. 76 00:05:55,430 --> 00:05:56,570 Now, that plot. 77 00:06:02,050 --> 00:06:03,370 We got a result what? 78 00:06:11,020 --> 00:06:12,310 In India, of course, the. 79 00:06:14,450 --> 00:06:18,440 We got the neural network and ascalon. 80 00:06:20,450 --> 00:06:22,940 I did not see the lie. 81 00:06:24,270 --> 00:06:28,280 So let me pause here and I will come back in a moment. 82 00:06:30,610 --> 00:06:38,220 So I did run, I was a secretary and as you say, there is a small line here, so and this one is smaller, 83 00:06:38,230 --> 00:06:40,390 it's basically nearly the same. 84 00:06:41,350 --> 00:06:42,310 So now that. 85 00:06:44,170 --> 00:06:52,910 Some call for the so before that, we need to analyze a dirham, so analyzing the diagram is clear that 86 00:06:52,970 --> 00:06:58,850 the current model shot across neural network model returned a better result. 87 00:06:59,650 --> 00:07:01,930 The poor are better suited to the lie. 88 00:07:02,080 --> 00:07:10,090 So the line here, they are divided equally in equal parts above and below and isolated by. 89 00:07:10,090 --> 00:07:13,150 I feel so isolated. 90 00:07:13,540 --> 00:07:22,300 If you're looking for this, we need to calculate our MSA for the linear regression model to do this 91 00:07:22,310 --> 00:07:26,210 without you Escalon matrix where our function. 92 00:07:27,040 --> 00:07:28,440 So let's try not. 93 00:07:30,250 --> 00:07:32,850 So I think the before I did. 94 00:07:37,650 --> 00:07:38,190 Go. 95 00:07:44,740 --> 00:07:50,590 So is it important for Escalon, Dot, Matrix? 96 00:07:51,830 --> 00:07:52,610 Import. 97 00:07:54,270 --> 00:07:59,280 Main square, main square, not absolute. 98 00:08:09,730 --> 00:08:14,080 And then we get you a Akko mean. 99 00:08:16,110 --> 00:08:16,700 Square. 100 00:08:17,820 --> 00:08:19,380 Mean squarer. 101 00:08:21,210 --> 00:08:21,750 Why? 102 00:08:25,680 --> 00:08:26,160 Has. 103 00:08:28,560 --> 00:08:29,670 Come, why? 104 00:08:31,110 --> 00:08:31,530 Brett. 105 00:08:33,170 --> 00:08:34,820 And then bring. 106 00:08:40,160 --> 00:08:40,940 Lenie. 107 00:08:42,990 --> 00:08:45,210 Regression model. 108 00:08:46,080 --> 00:08:47,630 And then pray. 109 00:08:50,100 --> 00:08:51,330 And as. 110 00:08:54,830 --> 00:08:59,030 Run the sale and we got zero point zero one five. 111 00:09:02,040 --> 00:09:05,140 So zero point zero one five one five. 112 00:09:05,640 --> 00:09:07,620 And we got before for a. 113 00:09:09,920 --> 00:09:12,220 Where is it going? 114 00:09:13,120 --> 00:09:14,190 The. 115 00:09:17,500 --> 00:09:20,770 Here, zero point zero. 116 00:09:23,460 --> 00:09:28,250 I'm sorry, I need you to copy this one so zero point zero six. 117 00:09:36,280 --> 00:09:40,570 So zero point zero six compared with zero point zero one five. 118 00:09:40,960 --> 00:09:44,190 So this one had a lower Mosquera. 119 00:09:44,230 --> 00:09:50,610 This one does mean darker than your network model is better than the linear regression model. 120 00:09:51,340 --> 00:09:56,600 So to get a better result, you train a neural network and you get a better result. 121 00:09:57,130 --> 00:09:59,310 And I is in this video. 122 00:09:59,680 --> 00:10:03,700 I hope you enjoy it and I hope you enjoyed our project. 123 00:10:03,910 --> 00:10:05,650 And I hope you learn a lot from it. 124 00:10:06,190 --> 00:10:11,950 So I will see you in another course with another project. 125 00:10:12,580 --> 00:10:13,240 Thank you.