1
00:00:00,120 --> 00:00:05,790
Halaal, before going ahead in this session, let's have a quick overview of what we have done in this

2
00:00:05,790 --> 00:00:06,230
project.

3
00:00:06,420 --> 00:00:12,020
So doing lots of things, feature engineering, feature and coding, all these types of things.

4
00:00:12,030 --> 00:00:18,600
In the last session, we have created our amazing machine learning model using cross-validation in which

5
00:00:18,600 --> 00:00:25,290
we have applied this logistic regression algorithm on our data and we have achieved that much accuracy

6
00:00:25,290 --> 00:00:26,100
on our data.

7
00:00:26,310 --> 00:00:31,920
So in this session, what we have to do, we have basically this assignment in which I have to apply

8
00:00:32,190 --> 00:00:36,730
multiple algorithms on our data and we have to check its accuracy as well.

9
00:00:37,080 --> 00:00:44,460
So what our algorithm, you know, whatever, you know what our what our good will execution no bias,

10
00:00:44,460 --> 00:00:52,020
let's say again in random forest decision, tree activist, whatever algorithm, you just implement

11
00:00:52,020 --> 00:00:53,130
it on this data.

12
00:00:53,550 --> 00:00:54,800
So this is exactly your task.

13
00:00:54,810 --> 00:00:58,340
So let me do you all these things in front of you.

14
00:00:58,560 --> 00:01:04,200
So I'm just going to say and if you guys don't know all these algorithms, you can check out my all

15
00:01:04,200 --> 00:01:11,220
the previous videos where I have explained each and every algorithm from scratch, including its mathematical

16
00:01:11,220 --> 00:01:17,220
intuition, because understanding the maths behind algorithm is very important, because whenever you

17
00:01:17,220 --> 00:01:23,790
are going to apply for a green job as a data scientist in any top notch companies, they are not going

18
00:01:23,790 --> 00:01:24,410
to ask who you are.

19
00:01:24,420 --> 00:01:26,940
What is the integration they are going to ask you.

20
00:01:26,940 --> 00:01:32,890
You are telling me how exactly you can achieve that best fit line in linear regression.

21
00:01:33,180 --> 00:01:39,010
So that's why you have to understand mathematics behind each and every algorithm.

22
00:01:39,060 --> 00:01:44,040
So in this session, what we have to do, we are basically going to consider navabi bias, logistic

23
00:01:44,220 --> 00:01:46,140
and random forest.

24
00:01:46,140 --> 00:01:47,220
And this is a tree.

25
00:01:47,490 --> 00:01:53,940
So we are basically going to cover five algorithms and all these five algorithms I have already covered

26
00:01:53,940 --> 00:01:56,250
in my all the previous lectures.

27
00:01:56,610 --> 00:01:59,110
Just go and check out all these things.

28
00:01:59,130 --> 00:02:01,070
So very first I have to import all these things.

29
00:02:01,080 --> 00:02:08,460
I'm going to say from this Escalon module I have something which is exactly my name by I just press

30
00:02:08,460 --> 00:02:08,820
tab.

31
00:02:08,820 --> 00:02:12,600
And from this I have to import my Gorshin and we offer it.

32
00:02:12,600 --> 00:02:13,230
What I have to do.

33
00:02:13,230 --> 00:02:19,680
I'm just going to say from this Escalon, I have to import something on this linear model.

34
00:02:19,680 --> 00:02:21,750
So I'm just going to send in an a school model.

35
00:02:21,750 --> 00:02:26,410
And from this I have to import my logistic regression with exactly this one.

36
00:02:26,730 --> 00:02:30,450
Now, what I have to do from this on, I have to import.

37
00:02:30,450 --> 00:02:31,380
This time I can.

38
00:02:31,380 --> 00:02:34,320
And so I'm going to see neighbors module.

39
00:02:34,320 --> 00:02:38,400
I have something known as my canibus.

40
00:02:38,400 --> 00:02:40,510
You can just press there.

41
00:02:41,040 --> 00:02:47,550
Here you have to import classifier because this is exactly your classification is now what you have

42
00:02:47,550 --> 00:02:55,350
to do from this side could learn you have something known as in Sambal Module and from this in Sambal

43
00:02:55,350 --> 00:03:02,700
module here you have done stuff in Zambia, learning algorithms like Random Forest actually boost and

44
00:03:02,700 --> 00:03:04,050
tons of algorithms.

45
00:03:04,380 --> 00:03:10,620
So I'm going to say I have to import my, let's say, random forest classifier for what I have to do.

46
00:03:10,620 --> 00:03:14,090
Let's say I have to import my let's say this is entry as well.

47
00:03:14,160 --> 00:03:21,630
So I'm just going to say from Sikandar Dautry, I have to import something known as decision tree classifier.

48
00:03:22,110 --> 00:03:24,330
Just aggregate all the stuff lexan.

49
00:03:24,360 --> 00:03:27,120
Just going to define a blank list over there.

50
00:03:27,390 --> 00:03:32,190
And in this blacklist I have to append each and every algorithm.

51
00:03:34,320 --> 00:03:38,190
Then once I have this list, then I'm going to trade on this list.

52
00:03:38,430 --> 00:03:40,440
That is exactly my approach over here.

53
00:03:40,770 --> 00:03:46,740
So I'm just going to say models dot a pen and what I have to open this time.

54
00:03:46,770 --> 00:03:52,200
Let's say I have to have been here in the form of Stupples, where I'm just going to say that if I have

55
00:03:52,200 --> 00:04:00,330
to append this logistic regression and basically this is exactly neme, I have to initialize my regression

56
00:04:00,330 --> 00:04:00,690
as well.

57
00:04:00,720 --> 00:04:07,830
So I'm just going to say logistic regression and just initialize it after what we have to do next.

58
00:04:07,920 --> 00:04:14,970
I'm just going to say I have to abide by another thing which is exactly this time I have to import Mangosteen

59
00:04:15,000 --> 00:04:19,850
and I'm just going to say this is exactly my like Sydney bias.

60
00:04:19,860 --> 00:04:22,680
I'm going to say this is my name bias.

61
00:04:23,040 --> 00:04:29,520
Once I have all these things, then I have to initialize its class as after doing all these things.

62
00:04:29,550 --> 00:04:33,790
Now what we have to do, we have to upend our whole case.

63
00:04:33,850 --> 00:04:38,270
So I'm just going to say third one is exactly my random forests.

64
00:04:38,270 --> 00:04:41,520
So I'm going to say exactly my random forest.

65
00:04:41,520 --> 00:04:43,390
Then I have to initialize its class as well.

66
00:04:43,980 --> 00:04:50,050
So this time I'm just going to say random forest classifier once having all the stuff.

67
00:04:50,080 --> 00:04:56,790
What I have to do, I'm just going to say this time I have to initialize or I have to pass my decision

68
00:04:56,790 --> 00:04:57,000
tree.

69
00:04:57,000 --> 00:05:00,320
So I'm just going to say this is exactly my decision tree.

70
00:05:00,330 --> 00:05:05,430
Then I have to initialize my decision tree as well, which is exactly classifier.

71
00:05:06,750 --> 00:05:13,500
After having all the stuff, you have to upend the algorithm as well, which is exactly OK in it.

72
00:05:13,680 --> 00:05:18,530
So I'm going to say this is my Ghanim and you have to initialize it as well.

73
00:05:18,540 --> 00:05:26,160
So I'm just going to say he niebler classifier and this time if you will pass shift plus that you will

74
00:05:26,160 --> 00:05:29,910
get all the parameters over here, you will get tons of parameters in this class.

75
00:05:30,480 --> 00:05:39,510
So now what you have to do, basically just execute the cell over there and now we have to fetch let's

76
00:05:39,550 --> 00:05:46,650
say let's say we have to print what exactly is my algorithm as well as we have to find all the stats

77
00:05:46,650 --> 00:05:48,300
of this particular algorithm.

78
00:05:48,310 --> 00:05:50,730
It means I just need two things for this.

79
00:05:50,730 --> 00:06:03,000
I'm just going to say for name comma models in model in this model list, once I have this iteration,

80
00:06:03,330 --> 00:06:06,240
after having this iteration, what do we have to do?

81
00:06:06,250 --> 00:06:10,470
Basically, I'm just going to say let's say I have to print my name.

82
00:06:10,470 --> 00:06:17,730
So I'm just going to say just print its name after it or whatever model I have on this.

83
00:06:17,730 --> 00:06:26,470
I'm just going to say just fit my X, underscore Krien and definitely Y underscore Grainne.

84
00:06:26,760 --> 00:06:29,580
Make sure you don't have any case sensitive issues.

85
00:06:29,850 --> 00:06:34,890
So here you will see your X is in top form and your guy is in a small one.

86
00:06:35,160 --> 00:06:39,020
So I'm going to say it is nothing but my X on a screen.

87
00:06:39,180 --> 00:06:42,150
So once you fit what you have to do, you have to do prediction.

88
00:06:42,150 --> 00:06:45,930
So I'm just going to say model dot predict.

89
00:06:46,020 --> 00:06:52,350
And here I have to say on X on the school test, I have to predict once doing prediction, I have to

90
00:06:52,370 --> 00:06:53,280
store somewhere else.

91
00:06:53,280 --> 00:06:59,970
Let's say I'm just going to say it is nothing but my let's say predictions at once having this prediction.

92
00:06:59,990 --> 00:07:05,100
Sorry, what I have to do, let's say I have to create my confused metrics.

93
00:07:05,400 --> 00:07:15,420
So let me just copy this code and let me just based there and let me just initialize my confusing metrics

94
00:07:15,600 --> 00:07:17,010
and hear what I have to say.

95
00:07:17,040 --> 00:07:22,110
I'm just going to say here, I have to say my predictions.

96
00:07:22,110 --> 00:07:23,220
The very first thing.

97
00:07:23,550 --> 00:07:27,840
The second thing is exactly might actually daviss my Y.A. test.

98
00:07:28,170 --> 00:07:31,380
Let's say I have to print this confusing metric.

99
00:07:31,380 --> 00:07:33,820
So I'm just going to print it after it.

100
00:07:33,840 --> 00:07:37,370
What we have to do, we have to print our accuracy as well.

101
00:07:37,380 --> 00:07:40,740
So here I'm going to say accuracy on the score is score.

102
00:07:41,190 --> 00:07:43,860
And the very first parameter is not predictions.

103
00:07:43,860 --> 00:07:48,060
What I have predicted and a second one, what is my actual data?

104
00:07:48,060 --> 00:07:50,060
Which is why NASCO test next.

105
00:07:50,090 --> 00:07:51,540
I have to print this as well.

106
00:07:51,540 --> 00:07:55,500
So I'm going to say just print all these stuff as well.

107
00:07:56,070 --> 00:08:03,200
Let's say after having all this stuff, let's say here I'm going to add my espacio this or that, or

108
00:08:03,210 --> 00:08:08,280
you can say adding my new learning model so that it will be more user friendly.

109
00:08:08,850 --> 00:08:13,380
So after having all this stuff, let me append on Lunine as well.

110
00:08:13,380 --> 00:08:16,850
So I'm going to say just append Newline over here.

111
00:08:16,860 --> 00:08:22,860
So if I'm going to execute this well now, you will see the very first thing is exactly my logistic

112
00:08:22,860 --> 00:08:23,490
regression.

113
00:08:23,490 --> 00:08:28,920
It has that much accuracy and it is exactly it is not my class model.

114
00:08:29,100 --> 00:08:33,150
And if you want to cross validated just about all these different.

115
00:08:33,150 --> 00:08:33,450
Different.

116
00:08:34,020 --> 00:08:40,440
The with respect to neighbors here, you will see this is exactly is confusing metrics, and this is

117
00:08:40,440 --> 00:08:45,520
exactly that accuracy with respect to this my neighbors as well.

118
00:08:45,730 --> 00:08:51,570
After that, you will see with respect to this random forest, you have a damn good accuracy.

119
00:08:51,570 --> 00:08:55,170
You will see your accuracy is 95 percent.

120
00:08:55,290 --> 00:08:59,580
And that's what I have defined in my random pilot's intuition.

121
00:09:00,040 --> 00:09:04,160
Random forest is exactly the best garden of data scientist.

122
00:09:04,650 --> 00:09:06,540
That's why this is exactly that reason.

123
00:09:06,540 --> 00:09:12,930
Whenever you have that much complex data and whenever you have some complex, massive data at the time,

124
00:09:12,930 --> 00:09:17,420
that random forest is going to be super, super over there.

125
00:09:17,790 --> 00:09:22,530
Similarly with the decision tree, it is still executing and it will take a while.

126
00:09:22,530 --> 00:09:23,220
Definitely.

127
00:09:23,220 --> 00:09:30,030
Now you will see with respect to this decision tree, you have somewhere close to your random security.

128
00:09:30,030 --> 00:09:39,210
It means whenever you are going to use it in Sambor learning techniques or new data, almost in 99 percent

129
00:09:39,210 --> 00:09:43,720
of the use cases, you will get your damn good accuracy.

130
00:09:43,980 --> 00:09:50,280
Similarly, with respect to your kenen, you have that definitely you have that much good accuracy.

131
00:09:50,790 --> 00:09:56,880
So you can simply say you are here whenever you have some pretty much complex use cases, your random

132
00:09:56,880 --> 00:10:04,770
forest, your decision tree, your booze, all your Sambell learning approaches are super, super pioneer

133
00:10:05,070 --> 00:10:05,360
there.

134
00:10:05,700 --> 00:10:07,230
So hope you love the session.

135
00:10:07,230 --> 00:10:13,620
And if you guys don't want to know all these algorithms, please go ahead and check out my all the previous

136
00:10:13,620 --> 00:10:20,460
videos in which I have explained each and every ill gotten from scratch, all these algorithms, all

137
00:10:20,460 --> 00:10:27,360
it's mathematical intuition, all these interna how exactly these algorithms works internally, because

138
00:10:27,360 --> 00:10:31,050
you have to make sure you know all these algorithms very well.

139
00:10:31,050 --> 00:10:33,120
So that's all about this project.

140
00:10:33,120 --> 00:10:35,100
Hope you love this project very much.

141
00:10:35,100 --> 00:10:35,860
Thank you.

142
00:10:35,880 --> 00:10:36,800
Have a nice day.

143
00:10:37,200 --> 00:10:37,900
Keep learning.

144
00:10:37,920 --> 00:10:38,850
Keep growing.

145
00:10:39,000 --> 00:10:40,020
Keep motivating.