1
00:00:00,150 --> 00:00:06,460
Helen, before going deep down to the session, let's have a walk through on all the previous sessions

2
00:00:06,490 --> 00:00:09,030
what we have done in this project till now.

3
00:00:09,390 --> 00:00:17,730
So from data in body to data analysis, by forming lots of data processing will also gain some meaningful

4
00:00:17,730 --> 00:00:18,850
insights from data.

5
00:00:18,870 --> 00:00:22,720
Let's say price was a source, price was the destination.

6
00:00:22,740 --> 00:00:30,210
How exactly prices for during what is the distribution of each and every metropolitan cities in terms

7
00:00:30,210 --> 00:00:30,730
of price?

8
00:00:31,050 --> 00:00:38,670
We are also faced these type of meaningful insights from data and we can simply represent to my stakeholders

9
00:00:38,710 --> 00:00:43,470
here, these are these these insights that I have extracted from data.

10
00:00:43,830 --> 00:00:50,910
And now after we have performed lots of techniques of encoding and all these different different estopinal

11
00:00:50,910 --> 00:00:54,540
also perform feature selection, outliers detection as well.

12
00:00:54,840 --> 00:01:01,110
So in this session, what we have to do, we have to apply a machine learning algorithm on our data

13
00:01:01,110 --> 00:01:04,470
as well as we have to automate on our prediction.

14
00:01:04,770 --> 00:01:07,920
What exactly is the meaning of this automation?

15
00:01:08,130 --> 00:01:13,680
So let's say I'm just going to define a function and one hour I will pass a machine learning algorithm

16
00:01:13,680 --> 00:01:14,570
to that function.

17
00:01:14,880 --> 00:01:21,510
All the stuff will be happen automatically by this function like, say, learning relationships from

18
00:01:21,510 --> 00:01:23,410
data check accuracy.

19
00:01:23,860 --> 00:01:24,930
What are the predictions?

20
00:01:25,110 --> 00:01:30,550
Checking the accuracy of model, dumping that model and all the different different like that things.

21
00:01:30,840 --> 00:01:36,900
So what I'm going to do very first, I just need some training data and some testing data so that I

22
00:01:36,900 --> 00:01:40,380
can create my model so that I can do some kind of prediction.

23
00:01:40,600 --> 00:01:46,890
So far, this what I'm going to do, I'm just going to say from this Ascalon library, I'm going to

24
00:01:46,890 --> 00:01:52,380
import already some module, which is exactly my train test is split.

25
00:01:52,590 --> 00:01:56,580
That will definitely helpful for you whenever you have to.

26
00:01:57,170 --> 00:02:02,400
You read any form of training and as what I have to do, I have to just execute it.

27
00:02:02,610 --> 00:02:05,190
And after it, we have to initialize this class.

28
00:02:05,400 --> 00:02:10,410
And if you will shift plus there you will get what exactly the documentation of function.

29
00:02:10,410 --> 00:02:14,020
You will see Taci train side and all the different different things.

30
00:02:14,340 --> 00:02:20,400
So very first here, you have to parse what is your independent data, what is a dependent data as well

31
00:02:20,400 --> 00:02:21,790
as what is your test side.

32
00:02:21,810 --> 00:02:30,190
So let's say I'm going to mention my test side or let's say I just need 20 percent of data as my testing.

33
00:02:30,420 --> 00:02:35,280
So here I'm going to say this and this side is equal to nothing but zero point two.

34
00:02:35,280 --> 00:02:38,100
It means my training suggests 80 percent.

35
00:02:38,110 --> 00:02:43,180
It means 80 percent of entire data will be considered for training purposes.

36
00:02:43,440 --> 00:02:50,940
So what it really does, it really don't meet acts on the train after it will be done on the score test.

37
00:02:51,150 --> 00:02:57,270
After that, it will hit a new Y in the score train as well as it will also return me.

38
00:02:58,140 --> 00:03:00,180
Why underscore test?

39
00:03:00,450 --> 00:03:05,300
So I'm just going to assign this and just execute all these stuff.

40
00:03:05,500 --> 00:03:07,460
Now what do we have to do?

41
00:03:07,590 --> 00:03:11,600
We have to basically apply machine learning algorithm on our data.

42
00:03:11,820 --> 00:03:14,040
So let's say I will define a function.

43
00:03:14,040 --> 00:03:16,680
Let's say its name is whatever you want.

44
00:03:16,920 --> 00:03:19,740
So Lexcen will define a function with name as pathetic.

45
00:03:19,920 --> 00:03:26,490
And here, whatever machine learning model I'm going to pass, it will give prediction with respect

46
00:03:26,490 --> 00:03:27,710
to that machine learning model.

47
00:03:27,930 --> 00:03:32,970
So very first, what we have to do, we have to learn some relations for our data supporters.

48
00:03:32,970 --> 00:03:37,440
I'm going to say what our machine learning model I'm going to pass on this.

49
00:03:37,440 --> 00:03:39,440
Using that, I have to call my friend.

50
00:03:39,720 --> 00:03:41,700
And here very first, I have to pass.

51
00:03:41,700 --> 00:03:44,690
My accent is called train and why in the school.

52
00:03:45,030 --> 00:03:46,830
So here I'm going to pass all this stuff.

53
00:03:47,070 --> 00:03:51,420
Let's say after learning, I'm going to store all the relationships in model.

54
00:03:51,630 --> 00:03:54,150
And here, that's a very first.

55
00:03:54,150 --> 00:03:56,370
I have to print something that's out of print.

56
00:03:56,610 --> 00:03:58,920
What exactly the training course for this?

57
00:03:58,920 --> 00:04:03,420
I'm going to say using this model, I have to call my something, which is score.

58
00:04:03,600 --> 00:04:07,380
And here I had to mention my accent is called train.

59
00:04:07,680 --> 00:04:14,220
And after that we have to mention a word is called let's say I have to print it in some user-friendliness

60
00:04:14,220 --> 00:04:14,460
here.

61
00:04:14,460 --> 00:04:17,850
I'm going to say I have to print it, let's say so here.

62
00:04:17,850 --> 00:04:20,760
I'm going to say nothing but my something.

63
00:04:20,760 --> 00:04:22,650
Let's say training school.

64
00:04:22,650 --> 00:04:29,580
So here I'm going to say it is my training is called and after let's say my training is called is something

65
00:04:29,730 --> 00:04:32,280
which is exactly placeholder over here.

66
00:04:32,490 --> 00:04:38,850
So whatever placeholder I have to find out what they're doing exactly the see values from my formal

67
00:04:38,850 --> 00:04:39,360
function.

68
00:04:39,600 --> 00:04:43,160
So it will receive values from this format function.

69
00:04:43,410 --> 00:04:48,720
So after what we have to do, let's say after it, we have to do prediction.

70
00:04:48,730 --> 00:04:53,070
So I'm going to say using my model, I have to use a private function.

71
00:04:53,310 --> 00:04:56,950
So basically we have to predict on our X test data.

72
00:04:57,360 --> 00:04:59,580
So after it, what we have to do, let's say I'm.

73
00:04:59,660 --> 00:05:07,260
Undistorted somewhere as time we to store it in, let's say, predictions in, which is exactly my error

74
00:05:07,660 --> 00:05:09,520
after what we have to do next.

75
00:05:09,550 --> 00:05:11,920
We have to protect our prediction.

76
00:05:12,190 --> 00:05:19,540
So for this, what we guys can do, you guys can say, let's say predictions are nothing, but predictions

77
00:05:19,540 --> 00:05:23,170
are going on and you can place a placeholder over here.

78
00:05:23,410 --> 00:05:28,940
And this placeholder will see values from your predictions.

79
00:05:29,800 --> 00:05:36,790
And after what we have to do, let's say I'm going to bring some Espace over there so that it will look

80
00:05:36,790 --> 00:05:37,990
something attractive.

81
00:05:37,990 --> 00:05:44,690
So I'm going to say just bring this place, let's say, offering all the stuff I have to check here.

82
00:05:44,920 --> 00:05:48,040
What exactly is my R-squared matrix?

83
00:05:48,280 --> 00:05:50,350
So that first, what do we have to do?

84
00:05:50,710 --> 00:05:51,410
Very flor's.

85
00:05:51,440 --> 00:05:58,810
I'm going to say from this Escalon, from this ascalon, I have to import something else, which is

86
00:05:58,810 --> 00:05:59,980
exactly matrix.

87
00:06:00,250 --> 00:06:10,180
And using this matrix, whatever I have imported Avel, I'm going to say this matrix dot r2 is just

88
00:06:10,190 --> 00:06:10,520
best.

89
00:06:11,020 --> 00:06:12,180
This is exactly right.

90
00:06:12,790 --> 00:06:19,470
And here you have to mention what is your actual data which is in lightest and what are your predictions,

91
00:06:19,480 --> 00:06:25,240
which is exactly in your predictions, which is what did and after let's say I have to store it somewhere

92
00:06:25,240 --> 00:06:29,110
else, let's say I'm going to say this is my let's say or two score.

93
00:06:29,140 --> 00:06:34,370
And after it, what we have to do, let's say I have to print this out as well.

94
00:06:34,390 --> 00:06:43,120
So here I'm going to say my R-squared value is nothing, but it is just this placeholder and my placeholder

95
00:06:43,360 --> 00:06:50,680
will receive values is exactly from this variable that I have defined or what, dear.

96
00:06:51,010 --> 00:06:55,540
So here I'm going to just mention this are two is Kaveri.

97
00:06:55,540 --> 00:06:57,240
What did after it?

98
00:06:57,250 --> 00:06:58,030
What we have to do?

99
00:06:58,030 --> 00:07:01,930
Let's say we have to print some more metrics for my regression.

100
00:07:02,440 --> 00:07:10,370
So what are those which are exactly my Amit mean absolute error and mzee meniscal at it and animacy,

101
00:07:10,390 --> 00:07:16,870
which is exactly what rwd me is that these are exactly three metrics in case of your regression use

102
00:07:16,870 --> 00:07:17,100
case.

103
00:07:17,350 --> 00:07:26,740
So for this I'm going to say my MFC, my and as he looks very first, I have to print my M80, so I'm

104
00:07:26,740 --> 00:07:31,870
going to say my M80 is nothing but it is nothing.

105
00:07:31,870 --> 00:07:42,910
But it is just my mat, Craig's dot and you have to access your mi underscore absolute underscore atter.

106
00:07:43,120 --> 00:07:49,300
And here you have to mention what is your actual data which is in this contest and what are your prediction

107
00:07:49,300 --> 00:07:51,290
which is exactly in your predictions.

108
00:07:51,610 --> 00:07:55,330
So after doing all this stuff, I'm just going to copy from here.

109
00:07:55,330 --> 00:07:57,190
Just paste, just paste.

110
00:07:57,400 --> 00:08:00,010
And this time I have to print for Matthei.

111
00:08:00,340 --> 00:08:04,780
This time I have to print exactly for Oremus.

112
00:08:04,870 --> 00:08:06,790
So I'm going to say this is Oremus.

113
00:08:07,240 --> 00:08:16,150
So this time I have to use another function, which is exactly I mean is quite error, which is exactly

114
00:08:16,510 --> 00:08:16,950
this one.

115
00:08:17,110 --> 00:08:17,490
Yeah.

116
00:08:17,770 --> 00:08:24,610
And this time I have to use a function which is nothing but which is just a root of which is just a

117
00:08:24,610 --> 00:08:25,730
root of this one.

118
00:08:25,990 --> 00:08:33,850
So for this what you guys can do, you guys can just call this number Bidart security function so that

119
00:08:33,850 --> 00:08:36,820
you will compute whatever message will return.

120
00:08:36,820 --> 00:08:39,550
It will just compute its current of this one.

121
00:08:39,610 --> 00:08:41,590
Let's say I have to visualize it.

122
00:08:41,590 --> 00:08:45,800
The are what is the difference between my predictions and my actual data.

123
00:08:46,000 --> 00:08:50,200
So for this, what I'm going to do, I'm going to say Ascendis dot this plot.

124
00:08:50,470 --> 00:08:56,050
And here very first to have to mention what is your actual data, which is not in my wildest and what

125
00:08:56,050 --> 00:08:59,980
is exactly your predictions, which is exactly in your predictions.

126
00:08:59,980 --> 00:09:06,520
Just like you let's say I have to apply random forest algorithm on data.

127
00:09:06,790 --> 00:09:14,020
It means you have to just pass this random forest class over there and it will do predictions.

128
00:09:14,020 --> 00:09:15,520
It will give you a training, a score.

129
00:09:15,520 --> 00:09:21,320
It will cover all the square metrics and all these different different types of metrics as well as this

130
00:09:21,390 --> 00:09:22,450
with as well.

131
00:09:23,020 --> 00:09:29,100
So what I'm going to do, I'm going to say from my Escalon module, I'm going to use that force.

132
00:09:29,110 --> 00:09:31,630
I have to import my sample.

133
00:09:31,630 --> 00:09:39,190
And in this example, I have my random forest agressor to just executed and now I have to just call

134
00:09:39,190 --> 00:09:40,450
my function.

135
00:09:40,630 --> 00:09:45,090
And here I have to just initialize this random forest degreaser.

136
00:09:45,130 --> 00:09:46,750
So just execute it.

137
00:09:46,750 --> 00:09:51,370
Now, you will see where all of this does gets executed with respect to training.

138
00:09:51,550 --> 00:09:57,730
You have ninety five percent of school and these are all your prediction with respect to different different

139
00:09:57,730 --> 00:09:58,400
indexes.

140
00:09:58,540 --> 00:09:59,470
This is our.

141
00:09:59,610 --> 00:10:04,380
Kolu, which say is, yeah, the model has 80 percent accuracy.

142
00:10:04,530 --> 00:10:08,430
After that, you will hear all the different different metrics as well.

143
00:10:08,460 --> 00:10:10,920
And this is exactly your distribution plan.

144
00:10:10,920 --> 00:10:18,570
You will see over here at the zero point, you have a B, it means more than more than almost 70 percent

145
00:10:18,570 --> 00:10:23,060
of data points are almost predicted in a right way.

146
00:10:23,250 --> 00:10:24,690
So they're in France.

147
00:10:24,840 --> 00:10:28,450
You can just watch from this distribution network.

148
00:10:28,560 --> 00:10:30,480
So that's all about the session.

149
00:10:30,840 --> 00:10:33,840
So the session will be very handy and very useful for you.

150
00:10:34,350 --> 00:10:35,610
That's all about guys.

151
00:10:35,940 --> 00:10:36,640
Thank you.

152
00:10:36,690 --> 00:10:37,590
Have a nice day.

153
00:10:37,650 --> 00:10:38,420
Keep learning.

154
00:10:38,430 --> 00:10:39,300
Keep growing.

155
00:10:39,540 --> 00:10:40,470
Keep practicing.