1
00:00:00,600 --> 00:00:07,650
All in this session, we are going to learn what exactly is this linear regression, which is exactly

2
00:00:07,660 --> 00:00:09,680
a regression algorithm.

3
00:00:10,170 --> 00:00:14,800
So you will see it is exactly a regression algorithm.

4
00:00:14,820 --> 00:00:21,180
So how how this algorithm works and what does that what does the mathematics behind this algorithm?

5
00:00:21,520 --> 00:00:28,440
We're also going to learn about all this stuff and what type of use use it in what type of practical

6
00:00:28,440 --> 00:00:31,830
use cases this this algorithm is going to.

7
00:00:31,830 --> 00:00:32,560
Very handy.

8
00:00:32,580 --> 00:00:33,090
Would you?

9
00:00:33,960 --> 00:00:41,670
So you have learned your school is something known as something you have learned in school X Y equals

10
00:00:41,670 --> 00:00:42,430
to Amex.

11
00:00:42,430 --> 00:00:50,010
Plus, you might have learned this in your I think when you are in class eight or maybe in your intermediate

12
00:00:50,010 --> 00:00:54,500
or maybe in your twenties too, this is what you have learned in your school.

13
00:00:54,780 --> 00:00:56,200
I'm just going to assume it.

14
00:00:56,550 --> 00:01:00,740
So what is there to see and what are all these things.

15
00:01:00,900 --> 00:01:01,320
So I.

16
00:01:01,580 --> 00:01:01,920
Nothing, nothing.

17
00:01:01,920 --> 00:01:03,690
What what is exactly.

18
00:01:03,690 --> 00:01:04,770
Stop and see.

19
00:01:04,800 --> 00:01:10,290
Just my constant and x x let's x is nothing.

20
00:01:10,290 --> 00:01:12,480
X just our datapoint.

21
00:01:12,960 --> 00:01:19,330
All you can see in terms of data it is nothing, but it is just a different feature.

22
00:01:20,100 --> 00:01:25,950
So with respect to this independent feature, you can see with respect to this independent feature,

23
00:01:26,130 --> 00:01:27,840
you have to predict something.

24
00:01:27,840 --> 00:01:34,070
And this is why which is exactly my dependent feature that you have to predict.

25
00:01:34,500 --> 00:01:35,010
That's it.

26
00:01:35,010 --> 00:01:37,300
That's all about my linear regression.

27
00:01:37,620 --> 00:01:45,360
Now, the question is, let's say when you have you let's say when X is given for how you have to find

28
00:01:45,360 --> 00:01:48,570
Y, how you have to find where the location is given.

29
00:01:48,570 --> 00:01:50,790
Let's say let's say you have some use.

30
00:01:51,060 --> 00:01:52,830
Let's say you have some practical use.

31
00:01:53,010 --> 00:02:00,720
Let's it on the let's say you have some data, let's say on the basis of oh, wait, let's say you have

32
00:02:00,720 --> 00:02:03,480
some Vitória Aleksi here.

33
00:02:03,480 --> 00:02:06,060
You have some height that you have to predict.

34
00:02:06,660 --> 00:02:11,220
Let's say you have some vague let's say in terms of centimeter, in terms of centimeters.

35
00:02:11,250 --> 00:02:17,480
Let's say you have some weight, one hundred one twenty, one forty one sixty and ended.

36
00:02:17,760 --> 00:02:25,130
And here this is this is exactly why this is exactly your height and this is your weight.

37
00:02:25,290 --> 00:02:26,120
Let me let me.

38
00:02:26,280 --> 00:02:27,590
It is all these things.

39
00:02:28,380 --> 00:02:29,410
So let me.

40
00:02:30,780 --> 00:02:35,700
So this is exactly your, let's say height and on the basis of height.

41
00:02:36,000 --> 00:02:37,590
So you have to predict fit.

42
00:02:38,500 --> 00:02:44,980
You can you can assume wastewater treatment as it's already, so let's say its weight is somewhere,

43
00:02:44,980 --> 00:02:50,230
let's say, in terms of cagy and which was you in terms of not not in terms of bone structure, which

44
00:02:50,230 --> 00:02:51,480
is only 58 or so.

45
00:02:51,490 --> 00:02:58,250
It has some 60 to somewhere, let's say 66 and it is 72 telltales.

46
00:02:58,300 --> 00:03:01,870
And so this is a data that is given to you.

47
00:03:02,160 --> 00:03:04,810
This is my data that is given to all you can.

48
00:03:04,810 --> 00:03:07,170
Tomatis athletic training data.

49
00:03:08,220 --> 00:03:14,850
And let's say in future hats, in future, you have to predict for what can be the weight of a person

50
00:03:14,850 --> 00:03:17,450
if a height of a person will be let's.

51
00:03:17,580 --> 00:03:19,350
One hundred eighty centimeter.

52
00:03:20,300 --> 00:03:25,760
So what you will do, you will find some kind of relationship from this data, you will find some kind

53
00:03:25,760 --> 00:03:33,380
of relationship in terms of why it goes to see how you have to find and how you have to find.

54
00:03:34,670 --> 00:03:36,590
So what is over there?

55
00:03:36,740 --> 00:03:44,060
Let me let me consider very simple values that because I think you let me just open a new page because

56
00:03:44,060 --> 00:03:47,810
you're on the basis of, let's say, years of experience.

57
00:03:49,100 --> 00:03:50,360
You have some felony.

58
00:03:52,550 --> 00:03:57,080
You will definitely achieve centrality of years of experience in this.

59
00:03:57,080 --> 00:03:59,470
So this is this this is this is this is this.

60
00:03:59,690 --> 00:04:04,670
So what what exactly is the meaning of this concept for this constant is nothing.

61
00:04:04,670 --> 00:04:08,870
But what what exactly is my salary?

62
00:04:09,080 --> 00:04:14,330
What exactly is my salary when my years of experience is zero.

63
00:04:14,330 --> 00:04:18,470
But I can say when I am a threshold, what exactly is my salary.

64
00:04:18,620 --> 00:04:19,100
That's it.

65
00:04:19,100 --> 00:04:20,610
That's all about my seat.

66
00:04:20,960 --> 00:04:23,210
So what you will do, you will put X zero over it.

67
00:04:23,360 --> 00:04:25,670
So this is the thing where my wife goes to see.

68
00:04:26,620 --> 00:04:35,050
So it means it means what exactly you salani, what exactly you salani what exactly is salary?

69
00:04:35,370 --> 00:04:40,360
Ranee you have a zero years of experience that a simple meaning of this.

70
00:04:40,600 --> 00:04:42,400
Why Alex policy.

71
00:04:42,400 --> 00:04:44,110
And here this is simple minimal.

72
00:04:44,110 --> 00:04:46,960
Seewald yet another question will arise.

73
00:04:47,110 --> 00:04:53,960
How to find this and how to find the C how to find this and how to find it.

74
00:04:53,990 --> 00:04:59,710
T. Let's say in a previous use because it will see, let's say you have to find what can we do it for

75
00:04:59,710 --> 00:05:05,730
apples and having that much height you have to find out how to find how to find it to date.

76
00:05:05,750 --> 00:05:08,090
The problem statement you can be selected.

77
00:05:08,170 --> 00:05:13,810
What I'm going to do, I'm just going to plot this data that I'm just going to plot this data and is

78
00:05:13,810 --> 00:05:17,950
going to say this is something my Y-axis, this is my legacy Access's.

79
00:05:18,190 --> 00:05:25,520
So on the basis of height, basically on the basis of height, you have to predict what can we go with.

80
00:05:25,780 --> 00:05:27,940
So I'm just going to zoom in on Y-axis.

81
00:05:27,960 --> 00:05:28,230
This is.

82
00:05:28,620 --> 00:05:30,120
This is my X here.

83
00:05:30,130 --> 00:05:30,950
You have some data.

84
00:05:31,550 --> 00:05:35,440
These are all my data points to do that, all my data.

85
00:05:35,440 --> 00:05:42,460
And let's say in future X in future, you have this data point, you have this data point for which

86
00:05:42,460 --> 00:05:45,690
you have to predict for which you have to predict.

87
00:05:47,260 --> 00:05:49,280
This is my data point for which I have to predict.

88
00:05:49,600 --> 00:05:56,380
So the equation that we will also, which is which we all have learned in our school, is basically

89
00:05:56,380 --> 00:05:57,680
y khoshjamal democracy.

90
00:05:57,880 --> 00:06:03,970
So if I'm going to draw a graph over here, so I will draw or I can see if I'm going to draw a line

91
00:06:03,970 --> 00:06:07,380
of this, which is nothing, but it is just like this one.

92
00:06:08,800 --> 00:06:10,090
It is just like this one.

93
00:06:10,120 --> 00:06:13,090
So this equation is nothing but viscose to see.

94
00:06:13,120 --> 00:06:18,280
That's what we all have learned about basic math, basic math, basically.

95
00:06:19,180 --> 00:06:22,330
So now you will think, what is this and what is this?

96
00:06:22,340 --> 00:06:23,440
And which is basically my.

97
00:06:24,550 --> 00:06:25,180
What is this?

98
00:06:25,180 --> 00:06:26,770
What exactly is this slope?

99
00:06:26,800 --> 00:06:28,300
What exactly is the slope?

100
00:06:28,480 --> 00:06:31,560
So this this this line will tell nothing.

101
00:06:31,570 --> 00:06:35,260
It is just this ratio, all this Y and X.

102
00:06:35,270 --> 00:06:41,470
So this is exactly why by X, which is nothing what which is just my slope, which is nothing.

103
00:06:41,470 --> 00:06:42,780
But we just my slope.

104
00:06:43,060 --> 00:06:45,310
So here I am going to say it is nothing but.

105
00:06:46,340 --> 00:06:54,350
The rate of change with respect to this X rate of change, with respect to X, so if you have learned,

106
00:06:54,350 --> 00:07:02,210
if you learn a bit of calculus, so this is nothing, but this gets written in such a very weird rate

107
00:07:02,240 --> 00:07:10,070
of Y rate, y reckons the rate of change of Y with respect to with respect to this X and this is this

108
00:07:10,070 --> 00:07:11,570
is what what Maisto.

109
00:07:12,540 --> 00:07:17,370
Or you can see this, it goes to 2010 as well here.

110
00:07:17,400 --> 00:07:20,940
This is exactly this is exactly my title here.

111
00:07:20,960 --> 00:07:21,870
This is exactly why.

112
00:07:21,870 --> 00:07:27,660
Teto so if I have to summarize this or I guess if I have to explain the slope it in layman's terms,

113
00:07:27,660 --> 00:07:29,790
then I can say this is nothing.

114
00:07:29,790 --> 00:07:32,080
But let's say what my intuition was.

115
00:07:32,100 --> 00:07:32,850
This is my letter.

116
00:07:32,880 --> 00:07:33,930
This is why.

117
00:07:33,930 --> 00:07:34,840
Karlstrom Explicit.

118
00:07:35,100 --> 00:07:36,770
So what is this Amoa here?

119
00:07:37,080 --> 00:07:39,810
If I if I will ask you, what is this over here?

120
00:07:39,960 --> 00:07:42,610
And you have to explain it easiest way.

121
00:07:42,810 --> 00:07:43,590
So it is nothing.

122
00:07:43,590 --> 00:07:54,810
But if if this X will increase or decrease by one unit, then what is a change in Y, that symbol meaning

123
00:07:54,810 --> 00:07:56,550
behind your slope.

124
00:07:56,560 --> 00:07:59,070
Or you can see at what factor.

125
00:07:59,070 --> 00:08:01,340
At what factor, at what factor.

126
00:08:01,770 --> 00:08:09,660
This Y will change if this X will increase by one unit or X will decrease by one unit.

127
00:08:09,900 --> 00:08:11,660
So let me consider some examples.

128
00:08:12,210 --> 00:08:14,400
Let's say this is why closer to X.

129
00:08:14,820 --> 00:08:18,410
So if I were to say Y by X, so did not want to buy one.

130
00:08:18,690 --> 00:08:29,250
So I'm going to simply say if X is going to increment by one unit, then Y gets increment by two unit.

131
00:08:30,590 --> 00:08:37,630
That that's a simple meaning behind this, and if I'm going to say why close to half of X, so if I

132
00:08:37,630 --> 00:08:44,090
will, I will say if if let's say if X gets increased by one unit over there.

133
00:08:44,470 --> 00:08:52,290
So if X get increased by one unit over here, so Y gets increased by just half unit, that's what that's

134
00:08:52,290 --> 00:08:53,740
what the meaning of the slope.

135
00:08:53,770 --> 00:08:59,050
So let's talk about some more perspective in which our slope come into existence.

136
00:08:59,470 --> 00:09:01,520
Selected Alexei.

137
00:09:01,570 --> 00:09:05,130
I have let's say you have to compute X, you have to compute.

138
00:09:05,140 --> 00:09:07,960
Let's say you have to compute slope between descent.

139
00:09:08,350 --> 00:09:10,230
You have to compute slope between this and this.

140
00:09:10,480 --> 00:09:15,730
So what you will do, basically, you will basically join both these points.

141
00:09:15,850 --> 00:09:20,410
You will basically join with this point and you will compute distance between this.

142
00:09:21,710 --> 00:09:28,250
This and distance between this, so this is nothing but just just Y, y, X letters, resistance's,

143
00:09:28,250 --> 00:09:31,960
let's say twenty four and the distance is 50.

144
00:09:32,360 --> 00:09:39,320
So here Mike Slope is nothing but twenty four by 15 in this case between between what.

145
00:09:39,320 --> 00:09:41,060
This point I had a slope.

146
00:09:42,310 --> 00:09:48,040
But what if that's it, but what if that's it, but what if you have to compute?

147
00:09:49,290 --> 00:09:52,770
You have to compute slope at a particular point.

148
00:09:52,980 --> 00:09:55,660
You have to compute slope at this particular point.

149
00:09:55,890 --> 00:09:59,170
So what can the slope in such scenarios for here?

150
00:09:59,310 --> 00:10:01,830
What we will do so let's say you would say no.

151
00:10:01,860 --> 00:10:08,070
Here I have a single point of how to compute, how I can compute slope to what what what you can what

152
00:10:08,070 --> 00:10:08,710
you can think.

153
00:10:08,730 --> 00:10:09,750
Let's see what you can think.

154
00:10:10,050 --> 00:10:13,030
Let me let me consider a point very close to this point.

155
00:10:13,030 --> 00:10:17,430
Let's say I would consider a very close point to this and then I will conclude this.

156
00:10:17,850 --> 00:10:23,460
So you will say that nothing what zettabytes, you know, up slope is not zero zero in such case.

157
00:10:23,670 --> 00:10:31,560
In such case, in such this slope is nothing but Delta Y, y, Delta X, which is nothing but which

158
00:10:31,560 --> 00:10:34,200
is nothing but be small change in Y.

159
00:10:35,200 --> 00:10:41,010
It's more genuine because if you are going to consider already brought us to this point very close,

160
00:10:41,010 --> 00:10:42,220
this point to this point.

161
00:10:42,410 --> 00:10:48,690
So definitely you have some Minuti, you have some minuti, and that is exactly a slope.

162
00:10:49,300 --> 00:10:50,410
So my slope is nothing.

163
00:10:50,410 --> 00:10:58,690
But I school is nothing but the statewide tax, which is nothing but a small change in Y, and it gets

164
00:10:58,690 --> 00:11:00,930
divided by this small change in X.

165
00:11:01,810 --> 00:11:05,200
So that is exactly your slope at a particular point.

166
00:11:05,260 --> 00:11:10,070
So in the upcoming session we are going to learn more about the regulation.

167
00:11:10,090 --> 00:11:16,360
We are also going to find what can be my best line or I can for my use case, because it is not necessary

168
00:11:16,360 --> 00:11:18,900
that this is a best practice deadline.

169
00:11:19,180 --> 00:11:23,700
So we have to find our best line as well, depending upon the Nagios.

170
00:11:23,710 --> 00:11:27,650
So in all of our upcoming session, we are going to deal with the scenario as well.

171
00:11:27,910 --> 00:11:28,680
So thank you.

172
00:11:28,690 --> 00:11:29,750
Have a nice day.

173
00:11:29,920 --> 00:11:30,940
Keep learning.

174
00:11:30,940 --> 00:11:31,810
Keep growing.

175
00:11:32,020 --> 00:11:32,890
Keep practicing.