1
00:00:01,410 --> 00:00:09,030
Hello, also in the session, we are going to learn what exactly is random forest and what use cases,

2
00:00:09,150 --> 00:00:14,370
what type of use cases, you can basically use this machine learning algorithm.

3
00:00:14,700 --> 00:00:24,690
So random forest is that type of algorithm which is used in case of classification as well as in case

4
00:00:24,690 --> 00:00:29,470
of classification as well as in case of regression as well.

5
00:00:30,000 --> 00:00:34,630
So you might have heard that random forest is a classification algorithm.

6
00:00:35,100 --> 00:00:41,850
Yeah, most of the cases it is used in classification, but it is not correct that it is a classification

7
00:00:41,850 --> 00:00:47,880
algorithm because inside could learn, which is a machine learning library, which is especially big

8
00:00:47,880 --> 00:00:49,170
for modeling purposes.

9
00:00:49,650 --> 00:00:58,920
In this cyclone, you have something known as random forest degressive class that is responsible for

10
00:00:58,940 --> 00:01:04,670
performing the Deanwood, or I guess it would deal with some type of regression use case.

11
00:01:04,920 --> 00:01:12,160
And we have a class known as random forest classifier, which is exactly for classification use cases.

12
00:01:12,630 --> 00:01:16,260
So what exactly is this random forest?

13
00:01:16,800 --> 00:01:18,210
So random forest is nothing.

14
00:01:18,450 --> 00:01:18,800
It is.

15
00:01:18,830 --> 00:01:24,690
It just follows it just follows my example learning approach.

16
00:01:25,610 --> 00:01:30,890
Random forest just follows in Sambell Learning approach.

17
00:01:31,300 --> 00:01:34,840
So what is this in Sambell Learning Approach?

18
00:01:35,440 --> 00:01:45,040
So Istanbul Learning Airport is all about basically it learns from multiple models, it learns from

19
00:01:45,370 --> 00:01:54,400
multiple models, and at the end it will combine all the learning and add that it will combine all the

20
00:01:54,460 --> 00:01:55,090
learning.

21
00:01:55,600 --> 00:02:00,190
That was my Istanbul approach will do.

22
00:02:00,400 --> 00:02:04,600
And Random Forest basically follows this type of approach.

23
00:02:05,260 --> 00:02:07,410
And random forest is nothing.

24
00:02:07,420 --> 00:02:10,060
It is just a classification.

25
00:02:10,060 --> 00:02:13,500
It is just a collection of multiple decision trees.

26
00:02:13,510 --> 00:02:18,210
It is just a collection of multiple decision trees.

27
00:02:18,520 --> 00:02:26,050
So before understanding this random forest, you have to understand what exactly is this decision?

28
00:02:26,050 --> 00:02:28,840
Trees, how you can build decision trees.

29
00:02:29,140 --> 00:02:32,530
What are the parameters on which using what?

30
00:02:32,680 --> 00:02:36,780
What are the parameters using that you can build your decision trees.

31
00:02:36,880 --> 00:02:38,260
You have to understand that.

32
00:02:38,470 --> 00:02:40,120
So let me open a new bit.

33
00:02:40,120 --> 00:02:44,170
And here I'm going to give you a quick overview.

34
00:02:44,170 --> 00:02:46,830
What is this is and trees and how you can build it.

35
00:02:46,870 --> 00:02:51,700
What is the background of this country's what are the parameters that you are going to play with once

36
00:02:51,700 --> 00:02:53,430
you try to build the trees?

37
00:02:53,890 --> 00:02:56,790
So what exactly is decision tree?

38
00:02:57,220 --> 00:02:58,840
What is this isn't.

39
00:02:59,410 --> 00:03:02,250
So this season tree basically it's.

40
00:03:02,510 --> 00:03:04,540
And then machine learning algorithm.

41
00:03:05,720 --> 00:03:11,870
For classification use cases as well as for regression use case as well.

42
00:03:12,710 --> 00:03:19,320
So this is that algorithm that is highly used in classification as well as in regression use case.

43
00:03:20,000 --> 00:03:23,120
So it is basically this decision tree.

44
00:03:23,420 --> 00:03:26,950
It is basically a base algorithm.

45
00:03:26,960 --> 00:03:34,760
I can see it is basically a base algorithm which is used in every conceivable technique, let's say,

46
00:03:35,210 --> 00:03:39,750
in case of random forest, in case of active use.

47
00:03:39,770 --> 00:03:42,130
So these are all those algorithms.

48
00:03:42,620 --> 00:03:46,330
These are all those algorithm that follows in Zambia learning.

49
00:03:46,530 --> 00:03:54,350
And in this in all this type of algorithm, basically my decision tree is heavily used because it is

50
00:03:54,350 --> 00:03:57,620
a base algorithm that is used in in Sambell Learning.

51
00:03:57,920 --> 00:03:59,740
So what exactly is this?

52
00:03:59,750 --> 00:04:02,700
This is and what is this is entry.

53
00:04:03,500 --> 00:04:10,610
So it is basically like a rule-based kind of thing, let's say, according to condition, according

54
00:04:10,610 --> 00:04:12,750
to condition, how you have to proceed.

55
00:04:13,130 --> 00:04:13,930
So that's it.

56
00:04:13,970 --> 00:04:16,160
That's all about why this is entry.

57
00:04:16,730 --> 00:04:23,040
So it is basically that type of algorithm where it will try to decide what I can say.

58
00:04:23,300 --> 00:04:27,640
It will try to take some decision or it will try to divide the data.

59
00:04:27,980 --> 00:04:28,540
Let's say.

60
00:04:28,670 --> 00:04:31,330
Let me consider a very basic example over here.

61
00:04:31,610 --> 00:04:32,860
Let me open a new page.

62
00:04:33,620 --> 00:04:34,040
Yeah.

63
00:04:34,340 --> 00:04:39,110
Let's see if you have to go for a party that would consider example.

64
00:04:39,620 --> 00:04:40,150
Consider.

65
00:04:41,570 --> 00:04:47,310
Let's say if you have to go for a party, then you have to do some season.

66
00:04:47,510 --> 00:04:49,140
So what can we make this season?

67
00:04:49,440 --> 00:04:52,280
Let's say I'm going to put a very basic condition.

68
00:04:52,290 --> 00:04:53,480
Let's see if.

69
00:04:54,670 --> 00:05:00,940
Time is, let's say, greater than 10 p.m. So in such case, I have to put a condition.

70
00:05:01,870 --> 00:05:07,060
So what if time is not between 10 p.m. and what if time is better than 10 p.m.?

71
00:05:07,330 --> 00:05:10,230
So in such case, if I'm going to say yes, it is, yes.

72
00:05:10,240 --> 00:05:12,320
So here I'm going to say I'm not going.

73
00:05:12,670 --> 00:05:14,000
So basically don't do.

74
00:05:14,170 --> 00:05:18,030
And if it is not, then I have another condition.

75
00:05:18,040 --> 00:05:18,520
Let's see.

76
00:05:19,300 --> 00:05:20,410
Do we have money?

77
00:05:22,230 --> 00:05:23,290
Do we have money?

78
00:05:23,460 --> 00:05:24,910
Do we have sufficient money?

79
00:05:25,350 --> 00:05:25,950
I can see.

80
00:05:26,520 --> 00:05:28,530
So then we have a skin condition.

81
00:05:28,860 --> 00:05:36,510
So if yes, if we if we have sufficient money, then definitely go go and enjoy.

82
00:05:38,460 --> 00:05:44,370
Similarly, I have another condition, what if I don't have to be sure the money, so in such case,

83
00:05:44,370 --> 00:05:45,570
I can also download?

84
00:05:46,650 --> 00:05:55,590
So in such case, I can also don't go for the major thing that we will notice over here is we are trying

85
00:05:55,590 --> 00:05:56,840
to divide data.

86
00:05:57,240 --> 00:06:00,930
We are basically tracked by data on the basis of two factors.

87
00:06:01,200 --> 00:06:06,940
The very first one is exactly this time, the second one on the basis of money.

88
00:06:07,290 --> 00:06:08,990
So let me let me down here.

89
00:06:09,570 --> 00:06:10,590
Let me go down here.

90
00:06:11,340 --> 00:06:16,560
So basically on the basis of time and on the basis of money.

91
00:06:17,590 --> 00:06:20,140
Basically, I have to take some decision.

92
00:06:21,240 --> 00:06:22,830
Whether yes or no.

93
00:06:23,870 --> 00:06:31,010
Whether years or so, let's say I'm going to put some random entries over here, let's see what if my

94
00:06:31,010 --> 00:06:31,940
time is 12?

95
00:06:32,060 --> 00:06:33,880
What if my time is 10?

96
00:06:34,010 --> 00:06:35,810
What is this nine?

97
00:06:35,820 --> 00:06:37,140
What is it?

98
00:06:37,400 --> 00:06:39,510
What if I have money?

99
00:06:39,550 --> 00:06:40,790
Yeah, what if I have money?

100
00:06:40,820 --> 00:06:42,050
What if I don't have money?

101
00:06:42,470 --> 00:06:43,880
What if I don't have money?

102
00:06:43,910 --> 00:06:44,410
What if.

103
00:06:44,720 --> 00:06:45,710
Yeah, I have money.

104
00:06:46,130 --> 00:06:48,020
So if I have this condition.

105
00:06:48,410 --> 00:06:54,620
If I have this condition here, basically here my decision will be no, because here you will see it

106
00:06:54,620 --> 00:06:55,660
is greater than 10.

107
00:06:56,060 --> 00:06:57,180
Similarly over here.

108
00:06:57,230 --> 00:06:59,420
I have no similar law here.

109
00:06:59,660 --> 00:07:02,030
I have no money over here.

110
00:07:02,090 --> 00:07:09,000
My decision will be yours because over here my time is less than ten as well as I have money as such.

111
00:07:09,010 --> 00:07:10,400
Excuse my decision would be yes.

112
00:07:11,060 --> 00:07:14,180
So you will see if I have to interpret this.

113
00:07:14,180 --> 00:07:19,880
I can see whenever my time is at 2:00 and I have money for an hour of my time is 8:00.

114
00:07:19,880 --> 00:07:24,580
It means I have to come in this brand and I have money that I have to come in this one.

115
00:07:25,010 --> 00:07:29,020
So it means I can go I can go for a party and enjoy.

116
00:07:29,480 --> 00:07:32,660
So basically what we are going to do in this.

117
00:07:32,690 --> 00:07:36,140
So basically we can have multiple features.

118
00:07:36,140 --> 00:07:36,890
In this case.

119
00:07:36,890 --> 00:07:41,140
We have this two features on the basis of the decision tree.

120
00:07:41,180 --> 00:07:43,010
I have basically two features.

121
00:07:43,010 --> 00:07:47,030
And on the basis of these two features, I have to take some decisions.

122
00:07:47,450 --> 00:07:51,350
Let's see what if we have a more number of features?

123
00:07:51,530 --> 00:07:53,030
So let me open a new bridge.

124
00:07:53,180 --> 00:07:56,420
What if we have a more number of features?

125
00:07:56,720 --> 00:08:01,670
Let's say let's say 50 or 100 Alexi's 60 odd thousand.

126
00:08:01,670 --> 00:08:03,380
Any number of features let's we have.

127
00:08:03,740 --> 00:08:08,390
So in such case, we have to draw a high radical tree like this one.

128
00:08:09,050 --> 00:08:10,100
So this is my bet.

129
00:08:10,100 --> 00:08:17,930
And that's if you will come over here, here, my parent node, you will see here, my parents node

130
00:08:17,930 --> 00:08:21,620
is time and this data node is exactly money.

131
00:08:22,040 --> 00:08:25,310
And this is my leave node where I have with this season.

132
00:08:26,030 --> 00:08:28,580
So similarly, you were here last year.

133
00:08:28,670 --> 00:08:30,160
This is my parents node.

134
00:08:30,170 --> 00:08:35,740
So this is exactly will be my parents node and these are exactly the same.

135
00:08:36,260 --> 00:08:40,100
This is my granddaughter, which is my data and what I can see.

136
00:08:41,250 --> 00:08:47,780
So these are all tunnels that contain some kind of information that contains some kind of data that's

137
00:08:47,790 --> 00:08:49,660
in some level.

138
00:08:49,680 --> 00:08:51,540
Over here, I can have this one.

139
00:08:51,840 --> 00:08:52,950
I can't have this one.

140
00:08:53,340 --> 00:08:54,720
I can have this one here.

141
00:08:54,720 --> 00:08:56,850
I have this one condition here.

142
00:08:56,850 --> 00:08:58,350
I have this one to condition.

143
00:08:58,800 --> 00:09:01,730
So let's get these all out of my life.

144
00:09:01,740 --> 00:09:02,040
No.

145
00:09:03,520 --> 00:09:11,310
So what what we will do, it will be like a hierarchy, this it will be hierarchically like this.

146
00:09:11,710 --> 00:09:18,690
So on the basis of some, on the basis on some condition, I have to basically do some decision.

147
00:09:18,850 --> 00:09:22,810
That's what my decision tree will do in this country.

148
00:09:23,230 --> 00:09:31,670
Whatever pattern we will learn from our training data, whatever pattern we will learn from our training

149
00:09:31,670 --> 00:09:33,370
data on which our.

150
00:09:34,430 --> 00:09:40,550
Relationship is going to establish basically what craniotomies on which our relationship is going to

151
00:09:40,550 --> 00:09:41,210
establish.

152
00:09:41,630 --> 00:09:50,240
So whatever we are going to learn from data, the similar pattern, the similar pattern, or I can see

153
00:09:50,240 --> 00:09:56,090
the similar hierarchy we are going to follow for our test data, the similar hierarchy we are going

154
00:09:56,090 --> 00:09:58,310
to follow for our prediction data.

155
00:09:58,790 --> 00:10:05,300
So let's say in a previous use case, which is exactly this one, let's say you have to do prediction

156
00:10:05,300 --> 00:10:10,640
for what if my time will be nine o'clock and what if I don't have sufficient?

157
00:10:10,960 --> 00:10:12,850
Next, you have to do prediction for this one.

158
00:10:13,200 --> 00:10:17,270
Let's say this is my unseen neglected this is my testing data.

159
00:10:17,420 --> 00:10:22,200
And in this case, this is exactly my training data.

160
00:10:22,520 --> 00:10:25,090
This is exactly my training data.

161
00:10:25,730 --> 00:10:33,270
So on the basis of this, Janita, you have this disease and you have this disease and this is entered

162
00:10:33,380 --> 00:10:39,590
DNG and basically by a lot of Volumnia, you have to do some kind of prediction.

163
00:10:39,840 --> 00:10:43,900
So basically, let's say you have to do prediction for this data.

164
00:10:44,120 --> 00:10:45,140
Let's say this data.

165
00:10:45,500 --> 00:10:54,050
So what we have what we have time as less than 10 p.m. it means I will come in this loop and now I'm

166
00:10:54,050 --> 00:10:57,160
going to say I don't have sufficient money to go for party.

167
00:10:57,500 --> 00:11:02,300
So after it, I will come I will come over here to here.

168
00:11:02,330 --> 00:11:04,990
My decision is I don't have to go.

169
00:11:05,270 --> 00:11:07,310
So my decision is exactly.

170
00:11:07,490 --> 00:11:08,650
I don't have to go.

171
00:11:08,810 --> 00:11:10,190
Let me open a new page.

172
00:11:10,200 --> 00:11:12,640
Let me this, this, this, this one.

173
00:11:12,800 --> 00:11:16,210
Yeah, you will see, this is my data.

174
00:11:16,340 --> 00:11:18,650
Let's say let's say I have this data.

175
00:11:19,190 --> 00:11:25,820
So let's say I have to create a hierarchy of this data or I can say I have to build such a tree so that

176
00:11:25,970 --> 00:11:30,710
in future, whenever I have some data so that I can do some kind of prediction.

177
00:11:30,890 --> 00:11:34,840
So let's say I have to be I have to build such a tree for this.

178
00:11:35,300 --> 00:11:43,610
So but a major factor over here is what feature I have to select as my parents or whether I have to

179
00:11:43,610 --> 00:11:46,070
select this time feature as my parents.

180
00:11:46,340 --> 00:11:50,090
And rest of all will be my subsequent node.

181
00:11:50,810 --> 00:11:57,520
And let's see whether money will be my node and let's say all other features will be my subsequent note.

182
00:11:57,650 --> 00:12:05,390
So how I have to select the question is that whether I have to select time as my parent or whether I

183
00:12:05,390 --> 00:12:08,580
have to select this money as my parent.

184
00:12:08,810 --> 00:12:11,720
So this is a major concern in this is a..

185
00:12:12,110 --> 00:12:17,480
This is a major concern in this country at the time of selecting better.

186
00:12:17,490 --> 00:12:20,060
Know how you have to select to parenthood.

187
00:12:20,330 --> 00:12:21,500
This is a major concern.

188
00:12:21,680 --> 00:12:29,900
So basically, we have to approach for this to select our parent or to build, because once we have

189
00:12:29,900 --> 00:12:32,450
a better node, we can easily build our decision.

190
00:12:32,720 --> 00:12:35,920
Let's say, what if we have, let's say, ten teachers?

191
00:12:36,080 --> 00:12:42,320
So once we have our parent, no, let's say once we have our parent node, then let's say we have another

192
00:12:42,320 --> 00:12:44,600
condition or another condition or whatever.

193
00:12:45,020 --> 00:12:53,240
Then the similar problem statement, the same problem statement I can again solve using again soil using

194
00:12:53,390 --> 00:12:55,970
X it how I have this better node.

195
00:12:56,210 --> 00:12:59,090
So let's say I have ten features over here.

196
00:12:59,090 --> 00:13:03,670
Let's say I have ten features already, assuming I have ten features over here.

197
00:13:03,950 --> 00:13:10,280
So basically using the concept of entropy, I will discuss later.

198
00:13:10,310 --> 00:13:12,400
I will give you a quick overview of this.

199
00:13:12,410 --> 00:13:20,680
So basically using concept of entropy and information gain, you can easily find out which can be better.

200
00:13:20,690 --> 00:13:20,960
No.

201
00:13:21,140 --> 00:13:26,270
So let's say who that will feature will have highest information.

202
00:13:26,270 --> 00:13:30,730
Gain can be my or I can see will be my Baranov.

203
00:13:30,740 --> 00:13:36,950
Let's say in this case, I have this time as the highest information gain.

204
00:13:37,220 --> 00:13:38,600
So I will do a better node.

205
00:13:39,530 --> 00:13:42,540
Let's see if I have multiple features over here.

206
00:13:42,850 --> 00:13:44,960
I have another feature as let's have three.

207
00:13:45,260 --> 00:13:50,510
So this following statement gets split it into this money and have three.

208
00:13:50,770 --> 00:13:53,870
Similarly, over here, I have some other attribute as well.

209
00:13:54,080 --> 00:14:00,050
So you will see again, you have this kind of column, Stutman, that you can solve using concept of

210
00:14:00,050 --> 00:14:02,270
entropy and information gain.

211
00:14:02,510 --> 00:14:10,280
Then again, you have some kind of tree and at the end you will be up having some kind of decision tree.

212
00:14:10,610 --> 00:14:14,980
So that's what my entropy and information gain look for.

213
00:14:14,990 --> 00:14:16,790
Let me open a new bed.

214
00:14:17,150 --> 00:14:19,610
So there are basically two approaches.

215
00:14:19,610 --> 00:14:21,710
There are basically two approaches.

216
00:14:22,340 --> 00:14:28,980
The very first one is exactly using concept of entropy and information gain.

217
00:14:29,450 --> 00:14:30,350
So using.

218
00:14:31,870 --> 00:14:41,400
This Entropia information gain, you will end up building some disease and similarly using my Guiney

219
00:14:41,410 --> 00:14:46,050
index, or you can also come as guinea impurity.

220
00:14:46,930 --> 00:14:54,040
So using this in Guinea impurity, you can also build up having some kind of vigilante because maybe

221
00:14:54,190 --> 00:15:02,350
there are some algorithm that uses as a genuine death or as a guinea index as the criterion to build

222
00:15:02,360 --> 00:15:06,220
these entries, whereas there are some algorithm that uses and probably.

223
00:15:07,240 --> 00:15:10,700
So it's all up to it's all up to you what exactly you want to use.

224
00:15:10,720 --> 00:15:14,760
So what exactly is this entropy information gain and all this stuff?

225
00:15:14,890 --> 00:15:21,580
So I'm going to give you a quick overview, because if I to start all these things, it will almost

226
00:15:21,580 --> 00:15:24,170
take out of classes to explain all this stuff.

227
00:15:24,490 --> 00:15:28,200
So what exactly is this information gain and all this kind of thing?

228
00:15:28,420 --> 00:15:35,440
So it is all about how random really does what is the probability of having randomness in your data?

229
00:15:35,650 --> 00:15:44,680
So basically what entropy will do, it will basically gives impurity present inside of our data and

230
00:15:44,680 --> 00:15:54,430
using a very basic formula with using a very basic formula, which is exactly this minus of B.I and

231
00:15:54,580 --> 00:16:04,330
log two of log two of by using this formula, using this formula, you can easily compute, you can

232
00:16:04,330 --> 00:16:11,170
easily compute the probability, you can easily compute the entropy of a particular feature where I

233
00:16:11,170 --> 00:16:15,670
is exactly the number of classes in that particular feature.

234
00:16:16,180 --> 00:16:22,090
So once you have entropy, once you have entropy, you have to compute information.

235
00:16:22,510 --> 00:16:25,260
So what exactly is this information gain.

236
00:16:25,510 --> 00:16:27,750
So information gain what exactly.

237
00:16:27,760 --> 00:16:36,910
It's basically based upon basically based upon entropy, based upon entropy feature.

238
00:16:37,420 --> 00:16:46,090
Which one is going to get our highest gain and let's say among Avron f2 f3 feature, if acto feature

239
00:16:46,090 --> 00:16:51,940
as the highest information gain, then we will select this feature as a parent node.

240
00:16:53,040 --> 00:17:01,380
So at the end, any future, whosever future has a high information gain, then that future will be

241
00:17:01,380 --> 00:17:03,570
selected as our paranoid.

242
00:17:04,020 --> 00:17:07,570
So what exactly is a basic formula behind this?

243
00:17:08,190 --> 00:17:13,900
So it is nothing but whatever entropy you have computed for a particular feature.

244
00:17:14,130 --> 00:17:15,400
What do you have to do now?

245
00:17:15,750 --> 00:17:29,700
It is basically one minus summation of S and of as in two E I will s and we add this as equals to total

246
00:17:29,700 --> 00:17:36,180
data point in that particular feature and then exactly out of total data points.

247
00:17:36,330 --> 00:17:39,780
How many we have all how many.

248
00:17:39,780 --> 00:17:42,420
We have data points in order.

249
00:17:42,660 --> 00:17:49,670
Let's say we have three classes in feature one that say we have three classes, let's zero one two in

250
00:17:49,680 --> 00:17:50,190
each other.

251
00:17:50,370 --> 00:17:56,430
So how many data points we have with respect to zero, how many with respect to one and how many with

252
00:17:56,430 --> 00:17:57,210
respect to do.

253
00:17:57,450 --> 00:17:59,590
That's what this will do.

254
00:17:59,940 --> 00:18:05,380
Similarly, what is this e e is nothing but what is the entropy of the zero?

255
00:18:05,490 --> 00:18:11,490
Similarly, in case of what is the entropy of one after it, I have to do just summation of this and

256
00:18:11,490 --> 00:18:14,330
it will give us it will return us some value.

257
00:18:14,640 --> 00:18:18,260
So let's say future two has highest information.

258
00:18:18,720 --> 00:18:27,240
So feature two will be my paranoid and let's say future one will become exactly like this.

259
00:18:28,510 --> 00:18:37,000
And now you have some kind of decision, so similarly, who's our future, who's our future has our

260
00:18:37,000 --> 00:18:42,910
highest information again, that that particular feature, will we get selected as a panel?

261
00:18:43,090 --> 00:18:45,580
That's what my information will do.

262
00:18:45,760 --> 00:18:53,740
And I have another factor, which is exactly Guiney building and using this fact that we can also build

263
00:18:53,740 --> 00:18:54,330
a precedent.

264
00:18:54,670 --> 00:18:59,890
So it's all up to you what you want to use, whether you have to come across with entropy or whether

265
00:18:59,890 --> 00:19:02,120
you have to come across to get the index.

266
00:19:02,500 --> 00:19:08,490
So what exactly is going you know, like we write down a very basic formula that it is exactly yours

267
00:19:08,500 --> 00:19:11,920
inside one building this this is entry.

268
00:19:12,490 --> 00:19:17,110
So it basically used one minus sigma of one.

269
00:19:17,110 --> 00:19:24,910
I goes to one to see where exactly number of classes in any feature, let's say in this feature one

270
00:19:24,910 --> 00:19:28,270
you have three classes, let's say zero one two.

271
00:19:28,630 --> 00:19:32,940
So here it is, nothing but B of I of squared.

272
00:19:33,370 --> 00:19:38,710
So what it will do basically it will basically do a summation of all these things.

273
00:19:38,710 --> 00:19:41,470
It will do summation or basically for B zero.

274
00:19:41,470 --> 00:19:46,870
What is the probability of this one or the probability of having one class for the burden of having

275
00:19:46,870 --> 00:19:50,200
two class and whatever the index will become?

276
00:19:50,560 --> 00:19:52,090
So who's your teacher?

277
00:19:52,090 --> 00:19:56,050
Who's or our future has the lowest Guiney index?

278
00:19:56,800 --> 00:20:03,540
Whosoever future has that lowest guinea index that will get selected as a betting board.

279
00:20:03,580 --> 00:20:05,410
So that's all about the session.

280
00:20:05,420 --> 00:20:07,210
Hopefully will love the session very much.

281
00:20:07,210 --> 00:20:10,630
How exactly I have summarized all these.

282
00:20:10,640 --> 00:20:13,560
This is a result of thinking the easiest to the.

283
00:20:13,930 --> 00:20:21,430
So basically in the next session I'm going to give you how exactly random forest works, how exactly

284
00:20:21,430 --> 00:20:28,870
it uses this is entry's inside and how random forest is able to do predictions on whether it's a regression

285
00:20:29,050 --> 00:20:31,390
case or whether it's a classification.

286
00:20:32,060 --> 00:20:34,000
So I hope you will love this session very much.

287
00:20:34,330 --> 00:20:34,910
Thank you.

288
00:20:35,080 --> 00:20:36,940
How nice to keep learning.

289
00:20:36,940 --> 00:20:38,740
Keep growing, keep practicing.