1
00:00:00,080 --> 00:00:05,580
Welcome back to the class of our course about the complete introduction to The Science Show in this

2
00:00:05,580 --> 00:00:13,230
class, we are going to talk about another algorithm that is really present inside of machine learning.

3
00:00:13,650 --> 00:00:16,770
And in this case, it would be logistic regression.

4
00:00:17,550 --> 00:00:23,640
So basically what we do today is we are going to work with another tool that is pretty important in

5
00:00:23,640 --> 00:00:25,890
machine learning that is called Eskild.

6
00:00:26,430 --> 00:00:33,810
Basically, Skillern is a machine learning library that will offer you various machine learning tools

7
00:00:33,810 --> 00:00:35,460
that you guys will be able to work with.

8
00:00:35,490 --> 00:00:38,130
So basically, you will have access to all the algorithms.

9
00:00:38,360 --> 00:00:41,010
So logistic regression will be in it.

10
00:00:41,580 --> 00:00:47,730
You will have all the other algorithms that we talked about a little bit before in the class in a more

11
00:00:47,730 --> 00:00:48,930
theoretical way.

12
00:00:50,220 --> 00:00:50,600
All right.

13
00:00:50,610 --> 00:00:54,750
So basically, the first thing that you guys will need to do is import.

14
00:00:54,750 --> 00:00:59,160
What import is killoran inside of your by expeditor?

15
00:00:59,610 --> 00:01:05,190
So that's always simply go on your file right here instead of your settings project and the operator,

16
00:01:05,610 --> 00:01:11,420
click on the plus button right here and simply write down ASCII learn.

17
00:01:13,320 --> 00:01:18,600
So when it's done, it should be this one right here, simply click on install package and you guys

18
00:01:18,600 --> 00:01:20,500
should be able to have it.

19
00:01:20,510 --> 00:01:23,090
So as you can see, I'm having it right here.

20
00:01:24,130 --> 00:01:24,350
Right.

21
00:01:24,390 --> 00:01:29,880
So when it's all done, the next thing that we want to do is really jump into it to jump into what we

22
00:01:29,880 --> 00:01:30,570
are going to create.

23
00:01:30,960 --> 00:01:37,080
So for the purpose of this class, what I decided to do to make it the most simple possible is to work

24
00:01:37,080 --> 00:01:42,560
with confusion, metrics, basically confusion, metrics that will give us well, we'll explain this

25
00:01:42,570 --> 00:01:50,880
more easily our model, and will allow us to know how well how our model is good in terms of predicting

26
00:01:50,880 --> 00:01:51,300
something.

27
00:01:51,510 --> 00:01:56,460
So basically, what I decided to do is create a really basic example.

28
00:01:57,150 --> 00:01:58,250
What I'm talking about example.

29
00:01:58,270 --> 00:02:03,540
What I mean here is let's say someone is looking for a job and this person have what?

30
00:02:03,540 --> 00:02:06,090
It's Emmonak management position job.

31
00:02:06,090 --> 00:02:08,640
And this person needs to have experience in three fields.

32
00:02:09,000 --> 00:02:13,710
And basically, depending on the experience of the person in those three fields, this person will get

33
00:02:13,710 --> 00:02:15,740
a job or not get a job.

34
00:02:15,750 --> 00:02:22,050
And basically, we want to predict the fact that this person will well, the accuracy of our system

35
00:02:22,050 --> 00:02:28,680
to predict if this person will get a job or not get a certain job depending of its experience.

36
00:02:29,640 --> 00:02:36,660
So what I'm planning to do is with pendas, create areas where we are going to store everything well,

37
00:02:36,750 --> 00:02:41,280
basically databases where we are going to create our own data.

38
00:02:41,470 --> 00:02:44,250
So basically you will create your own database in this case.

39
00:02:44,280 --> 00:02:49,980
So what I suggest you do is really use twenty five elements inside of your database.

40
00:02:50,400 --> 00:02:57,810
And from this moment, from this moment, we will create our little program that will be able to predict

41
00:02:57,810 --> 00:03:02,370
if yes or no, our model works fine or well, if it's accurate or not.

42
00:03:03,030 --> 00:03:03,840
So let's start.

43
00:03:04,320 --> 00:03:07,350
So basically the first thing that we need to do is pretty simple.

44
00:03:08,100 --> 00:03:14,950
We need to import the everything that we will need for our well, for our class.

45
00:03:15,540 --> 00:03:16,380
So pretty simple.

46
00:03:16,380 --> 00:03:18,360
We are going to start with Seabourne.

47
00:03:18,360 --> 00:03:24,220
So we'll import Seabourne, as I said.

48
00:03:25,230 --> 00:03:29,370
You also need matplotlib to be able to work with grafs.

49
00:03:30,600 --> 00:03:33,500
Well, live from that five block,

50
00:03:36,510 --> 00:03:39,170
then we'll need also would be Fender's.

51
00:03:44,430 --> 00:03:48,780
We also would need some things from Escalon.

52
00:03:48,960 --> 00:03:56,130
So basically we will import them from as she learned that model.

53
00:04:02,730 --> 00:04:05,790
We want to import train.

54
00:04:07,420 --> 00:04:16,750
This split and we also need logistic regression, since we are going to perform a logistic regression

55
00:04:16,750 --> 00:04:23,940
and show the results inside of a confusion matrix, so we are going to read it down right here.

56
00:04:40,120 --> 00:04:40,610
Here we go.

57
00:04:41,110 --> 00:04:45,560
And finally, the last thing that we will need just correct.

58
00:04:45,600 --> 00:04:48,150
To be sure, it's written correctly.

59
00:04:51,030 --> 00:04:58,170
Look, just the progression, here we go and finally, we'll need metrics so from.

60
00:04:59,910 --> 00:05:01,580
Escape, learn

61
00:05:04,750 --> 00:05:07,770
would need to import metrics.

62
00:05:09,510 --> 00:05:16,560
All right, so right now we have all the elements that we need to be able to create our well, basically

63
00:05:16,560 --> 00:05:18,090
our database.

64
00:05:18,600 --> 00:05:20,720
And so for the database, it's pretty simple.

65
00:05:20,730 --> 00:05:26,430
What I did is, as I said, I created twenty five elements or the elements are already pre created,

66
00:05:26,610 --> 00:05:27,780
so just show them to you.

67
00:05:28,530 --> 00:05:29,860
So basically it's pretty simple.

68
00:05:29,880 --> 00:05:35,680
I have my candidate right here or the person let's call it the person I prefer person.

69
00:05:37,740 --> 00:05:42,870
So basically the person will have experience in finances, management, logistics.

70
00:05:43,110 --> 00:05:46,500
And finally, do the person get work, yes or no?

71
00:05:47,220 --> 00:05:48,190
So it's pretty simple.

72
00:05:48,210 --> 00:05:54,660
Once again, we have our person, number one, and that what we have here is how many years of experience

73
00:05:54,660 --> 00:05:55,850
this person has.

74
00:05:55,860 --> 00:06:01,800
So they said that it could be one year or two years, five years, anything that you want.

75
00:06:02,250 --> 00:06:02,610
All right.

76
00:06:02,620 --> 00:06:07,410
So the first thing that you guys are going to do is create a simple database that looks something like

77
00:06:07,410 --> 00:06:07,680
this.

78
00:06:08,460 --> 00:06:15,480
And then what we'll do is we will simply put all this together and create a data frame where we are

79
00:06:15,480 --> 00:06:17,050
going to store everything.

80
00:06:17,820 --> 00:06:23,910
So when you guys are done, what will do will create is that a frame that we will call data?

81
00:06:27,420 --> 00:06:32,010
And inside of this data, we are going to with the help of vendors.

82
00:06:32,040 --> 00:06:34,110
So basically we are going to refer to pendas.

83
00:06:36,360 --> 00:06:41,490
We're going to create a data frame that will store person.

84
00:06:43,400 --> 00:06:49,790
So basically, all the elements inside of person and basically in our columns, we want to have finance,

85
00:06:49,790 --> 00:06:51,230
management and logistics.

86
00:06:55,030 --> 00:06:55,970
So here we go.

87
00:06:56,380 --> 00:06:59,830
So instead of our columns, we want to have finance

88
00:07:07,570 --> 00:07:10,030
management and.

89
00:07:13,590 --> 00:07:14,070
Here

90
00:07:17,370 --> 00:07:19,260
and finally, logistics.

91
00:07:27,370 --> 00:07:30,820
And what we need also is our get work.

92
00:07:36,230 --> 00:07:37,080
Here we go.

93
00:07:39,950 --> 00:07:42,140
Good work.

94
00:07:43,130 --> 00:07:49,280
All right, so right now we have everything that we need, the next thing we want to do is verify if

95
00:07:49,280 --> 00:07:50,270
everything works fine.

96
00:07:50,270 --> 00:07:52,910
So we'll simply print our little data.

97
00:07:53,060 --> 00:07:58,430
So we print our variable data just to be sure that everything works fine.

98
00:08:02,780 --> 00:08:07,170
All right, so as you can see, it's generating perfectly right here, everything works fine.

99
00:08:07,460 --> 00:08:08,980
So let me just explain the data to you.

100
00:08:09,000 --> 00:08:13,340
As I said here, we have our years of experience for the three, first of them.

101
00:08:13,340 --> 00:08:16,250
So finance, management and logistics and years of experience.

102
00:08:16,730 --> 00:08:21,200
And the next thing that we want to have, it would be zeros or ones.

103
00:08:21,440 --> 00:08:23,770
So basically, one means the person gets the job.

104
00:08:23,780 --> 00:08:26,430
Zero means the person doesn't get the job.

105
00:08:26,810 --> 00:08:28,730
So it's pretty simple to understand.

106
00:08:29,750 --> 00:08:30,100
All right.

107
00:08:31,160 --> 00:08:36,140
So right now that we have everything that works here, it's time to get to the next step, which could

108
00:08:36,140 --> 00:08:39,120
be creating our logistic regression in Python.

109
00:08:40,070 --> 00:08:44,750
So for our logistic regression, what we'll need is a pretty simple we'll need two variables that will

110
00:08:44,750 --> 00:08:45,790
be X and Y.

111
00:08:45,800 --> 00:08:48,440
So basically they will be inside of our graph.

112
00:08:49,040 --> 00:08:55,910
So for our variable X, what we will have is pretty simple, will have the elements, finance, management

113
00:08:55,910 --> 00:08:57,750
and logistic as well.

114
00:08:58,370 --> 00:09:01,870
So what we'll do right now, we'll simply make a reference to person.

115
00:09:02,450 --> 00:09:13,130
So write down a variable person and from person we want to use, in this case, finance.

116
00:09:19,810 --> 00:09:20,570
Management

117
00:09:24,490 --> 00:09:27,430
and finally, logistics.

118
00:09:32,190 --> 00:09:38,510
The second variable that will have to be the variable, why and this let's put it in small letters and

119
00:09:38,520 --> 00:09:40,590
this variable will get worse.

120
00:09:40,660 --> 00:09:42,270
Basically, once again from person.

121
00:09:44,210 --> 00:09:51,080
We want to have get rid of the person, gets work for no right.

122
00:09:51,140 --> 00:09:53,410
So right now we have our model.

123
00:09:53,840 --> 00:10:01,110
The next thing would be creating our test, using the test split function to be able to test everything.

124
00:10:01,130 --> 00:10:06,830
So basically, we are going to decide that what percentage of the database will be used for testing

125
00:10:07,280 --> 00:10:10,400
and which one will be used for training.

126
00:10:11,660 --> 00:10:12,500
So pretty simple.

127
00:10:12,510 --> 00:10:13,760
How exactly do we write this?

128
00:10:14,690 --> 00:10:16,370
Simply write down X.

129
00:10:16,880 --> 00:10:18,100
So put it right here.

130
00:10:20,170 --> 00:10:26,020
X train X this.

131
00:10:28,920 --> 00:10:34,320
Then why train and why this?

132
00:10:37,480 --> 00:10:48,730
All of this would be equal to train this split is what I'm writing right now, is what?

133
00:10:49,000 --> 00:10:56,320
Well, how much I want for my data, how much I want of my data set to be used to test and what percentage

134
00:10:56,320 --> 00:10:59,890
of my data set I want to be to be used for training purposes.

135
00:11:00,610 --> 00:11:03,490
So basically, I'll finish writing and I'll explain everything.

136
00:11:04,210 --> 00:11:08,800
I love X Y, so we'll have our test sites.

137
00:11:08,800 --> 00:11:13,260
So in our case, we want a test site of, let's say, 30 percent.

138
00:11:13,630 --> 00:11:16,420
So this means that zero the three.

139
00:11:16,840 --> 00:11:20,950
So this means that we want what we want to use.

140
00:11:20,980 --> 00:11:28,060
Thirty percent of our data set for testing purposes and the other 70 percent will be used for, well,

141
00:11:28,060 --> 00:11:28,810
training purpose.

142
00:11:30,550 --> 00:11:36,970
And the second one will be random state and this will be equal to zero.

143
00:11:39,050 --> 00:11:44,950
All right, so when it's done so right now, we with this line of code right here, we are able to say

144
00:11:45,180 --> 00:11:46,550
what will be the test split?

145
00:11:46,550 --> 00:11:52,710
So how much would be used for testing purpose and how much would be used for training purposes?

146
00:11:52,710 --> 00:11:58,070
So we have everything that we need right there, right when it's all done.

147
00:11:58,070 --> 00:12:03,310
The next thing is pretty simple is writing our logistic regression model.

148
00:12:03,320 --> 00:12:04,190
So it's pretty simple.

149
00:12:04,190 --> 00:12:05,990
We'll do it down there.

150
00:12:08,530 --> 00:12:11,780
So it's going to be logistic regression.

151
00:12:13,520 --> 00:12:20,000
So in our case, let's call it yeah, let's call it logistic Craig, which would be pretty simple and

152
00:12:20,000 --> 00:12:20,440
logistic.

153
00:12:20,440 --> 00:12:27,650
Craig would be equal to logistic regression since we are applying logistic regression to it.

154
00:12:29,390 --> 00:12:33,810
So this would be available in which we are going to store the functional logistic regression.

155
00:12:36,330 --> 00:12:37,070
All right.

156
00:12:37,220 --> 00:12:39,470
Then the next thing is pretty simple.

157
00:12:39,650 --> 00:12:41,820
We'll call it let's call it L.G., L.R..

158
00:12:42,440 --> 00:12:45,640
It's going to be smaller and more easier to work with.

159
00:12:47,090 --> 00:12:51,800
So we'll have to r, which would be equal to logistic regression, will not have to write down each

160
00:12:51,800 --> 00:12:52,990
time logistic regression.

161
00:12:54,170 --> 00:13:00,410
So we'll have are that fit since we're fitting right now our training of X and Y.

162
00:13:03,660 --> 00:13:08,950
So X thing as well as Y train.

163
00:13:11,610 --> 00:13:14,190
The banks and capital letters.

164
00:13:16,530 --> 00:13:19,050
All right, so right now we have everything that we need.

165
00:13:23,030 --> 00:13:24,950
Real important, not making mistakes.

166
00:13:26,840 --> 00:13:33,090
Right, next thing that we need is we want to predict our well, our question right here.

167
00:13:33,620 --> 00:13:34,310
So why?

168
00:13:37,080 --> 00:13:38,940
And in this case, we'll call it prediction.

169
00:13:42,070 --> 00:13:43,600
We'll be equal to.

170
00:13:45,980 --> 00:13:49,130
In our case, logistic.

171
00:13:55,870 --> 00:14:04,850
That got to so equal to or not that and what I do want to predict the test of X, so in this case,

172
00:14:04,850 --> 00:14:09,260
predict X that test.

173
00:14:10,310 --> 00:14:15,220
So basically right now we have everything that we need to be able to run our model.

174
00:14:16,520 --> 00:14:21,710
Then the next thing that we want to do, since we want to put everything inside of a confusion matrix,

175
00:14:21,710 --> 00:14:25,280
we will create our confusion matrix right now.

176
00:14:25,290 --> 00:14:32,220
So basically our confusion matrix, what will be created with the use of once again, pendas.

177
00:14:33,320 --> 00:14:34,130
So pretty simple.

178
00:14:34,130 --> 00:14:36,310
How exactly do we create our confusion metrics?

179
00:14:36,950 --> 00:14:42,680
We are at first to create our variable, which would be called the math.

180
00:14:43,460 --> 00:14:44,100
So let's call it.

181
00:14:46,280 --> 00:14:50,510
Matt, let's use small letters as well.

182
00:14:55,800 --> 00:15:00,870
All right, so in this case, we'll have confusion, matrix and the confusion matrix would be equal

183
00:15:00,870 --> 00:15:01,950
to in this case.

184
00:15:09,900 --> 00:15:16,230
So we'll have our white test as well as our white prediction.

185
00:15:20,900 --> 00:15:22,710
Then we'll have padrone.

186
00:15:22,730 --> 00:15:28,910
So how how are things in south of our conclusion that in this part of the code we're creating our confusion

187
00:15:28,910 --> 00:15:29,450
metrics?

188
00:15:30,020 --> 00:15:31,220
And as I said.

189
00:15:33,820 --> 00:15:34,400
Here we go.

190
00:15:37,320 --> 00:15:43,400
So right here, as I said, what we'll do is we will give real names to our well, to our heroes.

191
00:15:43,730 --> 00:15:44,440
So pretty simple.

192
00:15:44,460 --> 00:15:46,440
Let's do it right here.

193
00:15:49,700 --> 00:15:52,190
So real names have real

194
00:15:54,800 --> 00:16:02,100
names, so for the first row, we will have right there, let's call it.

195
00:16:03,440 --> 00:16:05,390
I don't know how we can call it, we can call it.

196
00:16:05,390 --> 00:16:07,240
So it's right now.

197
00:16:07,790 --> 00:16:09,140
So basically true results.

198
00:16:09,650 --> 00:16:10,370
So true.

199
00:16:14,390 --> 00:16:20,620
And four columns, since it's everything that will be predicted, let's call it, I don't know.

200
00:16:26,050 --> 00:16:32,430
Let's call it provisions, since this is what will be predicted, so we'll call it a revision.

201
00:16:33,700 --> 00:16:34,090
All right.

202
00:16:34,480 --> 00:16:35,880
So we have everything that we need.

203
00:16:35,890 --> 00:16:42,880
The next thing that we need to do is we are going to use Seabourne to create our confusion metrics right

204
00:16:42,880 --> 00:16:43,070
now.

205
00:16:43,090 --> 00:16:47,430
So pretty simple, isn't that heat map?

206
00:16:47,440 --> 00:16:53,020
So basically, in this case, we're going to use a map to be able to see everything well, to be able

207
00:16:53,020 --> 00:16:56,770
to see it more clearly when our confusion metrics will be generated.

208
00:16:58,170 --> 00:17:03,660
And in this case, we are going to apply our map to our confusion magic, so you'll see it's going to

209
00:17:03,660 --> 00:17:07,680
be something that is pretty cool in this case.

210
00:17:07,680 --> 00:17:09,180
We call it math.

211
00:17:14,980 --> 00:17:19,190
All right, so we have everything that we need and annotation is going to be true.

212
00:17:21,580 --> 00:17:21,910
All right.

213
00:17:21,920 --> 00:17:23,930
So we have everything that we need.

214
00:17:24,520 --> 00:17:28,950
And finally, the last step will simply be printing everything.

215
00:17:28,960 --> 00:17:30,750
So basically, we want to print absolutely everything.

216
00:17:30,760 --> 00:17:31,270
So print.

217
00:17:35,920 --> 00:17:36,670
Actressy.

218
00:17:52,400 --> 00:17:54,890
We hear what happened are accuracy score.

219
00:18:18,110 --> 00:18:24,620
Here will have our white test, so we'll have our white test and that what we have predicted for our

220
00:18:24,980 --> 00:18:26,250
white as well.

221
00:18:28,130 --> 00:18:37,280
So here for our extra square, we'll have our white test, as well as what we have predicted to our

222
00:18:37,280 --> 00:18:44,470
White House for training and predict the battles that are right here.

223
00:18:50,860 --> 00:18:52,680
That's going to be why this.

224
00:18:58,430 --> 00:18:59,900
Also here, we'll have our.

225
00:19:02,450 --> 00:19:07,340
Why train as well as our why?

226
00:19:11,330 --> 00:19:11,930
Predicted.

227
00:19:18,060 --> 00:19:20,350
Here right now, we have everything that we need.

228
00:19:25,900 --> 00:19:32,470
So very important, the really important part of what we have done, doing all this would be correcting

229
00:19:32,530 --> 00:19:39,190
our code, because usually when we write down a huge code like this, it's really important to correct

230
00:19:39,190 --> 00:19:41,870
absolutely everything that we have just done.

231
00:19:42,640 --> 00:19:42,990
All right.

232
00:19:43,000 --> 00:19:46,300
So the last step will simply be writing down Kielty.

233
00:19:47,200 --> 00:19:55,690
That should be able to see everything that we have to do right now is we are going to run everything

234
00:19:55,690 --> 00:19:57,070
to see if it works perfectly.

235
00:19:57,070 --> 00:20:01,040
And if not, we're simply going to correct mistakes inside of our code.

236
00:20:01,060 --> 00:20:02,470
So if we run everything.

237
00:20:04,100 --> 00:20:09,980
OK, we can see that there is a mistake, so zero size error reduction operation if men, which has

238
00:20:09,980 --> 00:20:17,570
not identified so what we'll do right now, we'll simply look at our code and try to find the mistakes

239
00:20:17,570 --> 00:20:18,770
that are inside of it.

240
00:20:18,830 --> 00:20:22,220
What could be the mistakes that usually happen?

241
00:20:22,730 --> 00:20:27,210
So usually the mistakes that happen in those types of code will be basic mistakes.

242
00:20:27,230 --> 00:20:29,760
So I already see some of them.

243
00:20:29,780 --> 00:20:31,670
So basically I can see it right here.

244
00:20:31,680 --> 00:20:32,980
So we have whitelist.

245
00:20:33,380 --> 00:20:37,480
There should be well, we should have wide prediction that is right here.

246
00:20:37,490 --> 00:20:39,060
So we have this mistake right there.

247
00:20:39,590 --> 00:20:44,640
OK, so basically the mistake that we have is pretty simple.

248
00:20:44,960 --> 00:20:47,560
So we have whitest white men in here.

249
00:20:47,570 --> 00:20:50,660
We should have white prediction instead of white trim.

250
00:20:52,430 --> 00:20:54,620
So here will break down what prediction?

251
00:20:56,420 --> 00:20:57,290
And finally.

252
00:20:58,310 --> 00:21:01,790
The other place where we should have wide prediction should be here.

253
00:21:02,120 --> 00:21:03,760
So this is exactly what we will do.

254
00:21:05,880 --> 00:21:14,280
So we'll keep our white list and we are going to write down our white prediction.

255
00:21:16,220 --> 00:21:16,580
All right.

256
00:21:16,590 --> 00:21:19,540
So normally right now, we should have everything that we need.

257
00:21:20,030 --> 00:21:27,260
And if you guys if you guys you have an error inside of your code or whatever, you can simply copy

258
00:21:27,530 --> 00:21:28,970
my code right here.

259
00:21:29,000 --> 00:21:30,340
It should work properly.

260
00:21:31,580 --> 00:21:36,830
All right, so at the end, this is the result that you guys should receive, and this is a perfect

261
00:21:36,830 --> 00:21:39,360
result of usually what you guys showed up.

262
00:21:39,560 --> 00:21:41,060
So here you have your accuracy.

263
00:21:41,060 --> 00:21:44,690
So basically the accuracy of our model is 75 percent.

264
00:21:44,700 --> 00:21:51,230
So we are able to predict at 75 percent how many people will be well, how many people will get the

265
00:21:51,230 --> 00:21:54,320
job and how many people won't get the job.

266
00:21:55,910 --> 00:21:58,030
How exactly does it predict?

267
00:21:58,100 --> 00:21:58,640
Seventy five.

268
00:21:58,640 --> 00:22:00,110
So it's pretty simple.

269
00:22:00,110 --> 00:22:03,230
The calculation of this is the truth.

270
00:22:03,230 --> 00:22:10,580
So basically, what are the what, what the what is the truth of basically the truth that we can see

271
00:22:10,580 --> 00:22:13,550
it this way us and versus the provisions.

272
00:22:14,360 --> 00:22:17,810
So basically here will have our provisions for this provision of a Yes.

273
00:22:17,810 --> 00:22:18,660
Provision of a no.

274
00:22:19,010 --> 00:22:20,420
So here we'll have our true.

275
00:22:20,420 --> 00:22:20,870
Yes.

276
00:22:21,140 --> 00:22:22,700
A true right here.

277
00:22:22,940 --> 00:22:26,690
And basically how we make this calculation, it's pretty simple.

278
00:22:26,700 --> 00:22:33,440
So we are going to make an addition of what are all the provisions that we have made that happened.

279
00:22:33,440 --> 00:22:38,330
So basically, it's going to be this one right here and this one right here versus all the provisions

280
00:22:38,330 --> 00:22:40,040
that we made that basically didn't happen.

281
00:22:40,040 --> 00:22:43,880
So basically, we made mistakes right here and zero mistakes right there.

282
00:22:44,270 --> 00:22:47,630
So it's simply going to be eight plus one.

283
00:22:48,510 --> 00:22:55,400
And this divided by the number of provisions, which would give us 75 percent, which is which makes

284
00:22:55,910 --> 00:22:57,320
pretty much a lot of sense.

285
00:22:58,370 --> 00:23:05,450
So we can see this is how we are performing a a well, this type of regression.

286
00:23:06,170 --> 00:23:06,950
So you can see it's not.

287
00:23:07,200 --> 00:23:09,050
Well, yes, it could be complicated.

288
00:23:09,050 --> 00:23:14,940
But at the end of the day, your goal is really to have this to see how accurate is your model.

289
00:23:15,290 --> 00:23:20,240
So in our case, it's simply going to be nine divided by 12, since we are going to make eight plus

290
00:23:20,240 --> 00:23:23,950
one divided by the well of this.

291
00:23:23,950 --> 00:23:29,120
So basically eight plus three plus one that should give us seventy five percent and seventy five percent

292
00:23:29,120 --> 00:23:32,660
of accuracy is pretty good for this type of model.

293
00:23:33,740 --> 00:23:36,980
So as you can see, this would be the first model that we will create right now.

294
00:23:37,010 --> 00:23:38,770
So the first machine learning model.

295
00:23:38,780 --> 00:23:40,450
So this is the code around it.

296
00:23:40,460 --> 00:23:46,280
So basically I try to make it as simple as possible, but it's pretty normal that you guys have mistakes

297
00:23:46,280 --> 00:23:48,890
in a code that is big like this.

298
00:23:50,760 --> 00:23:54,020
But once again, if you guys still have mistakes, pretty simple.

299
00:23:54,020 --> 00:24:00,570
Just screenshot my code and you can simply retype it and you should have everything that you need.

300
00:24:00,950 --> 00:24:04,400
So that's a first class guys insu out in our next class.