1
00:00:00,120 --> 00:00:00,380
Hello.

2
00:00:00,720 --> 00:00:06,720
So before going ahead in this session, let's have a walkthrough on all our previous session from data

3
00:00:06,720 --> 00:00:13,620
importing by doing law to the analysis, by doing lots of data processing on our data after we have

4
00:00:13,620 --> 00:00:20,880
performed lots of speech and coding as well as our detection separating our independent features, feature

5
00:00:20,880 --> 00:00:24,490
selection, lots of ping we have done in this project.

6
00:00:24,750 --> 00:00:31,620
So in this session, what we have to do in the previous session, whatever automation we have done with

7
00:00:31,620 --> 00:00:37,020
respect to this automation, we have to play with our multiple algorithms.

8
00:00:37,050 --> 00:00:39,750
This is exactly the assignment for the station.

9
00:00:39,990 --> 00:00:45,030
And the very first statement is you have to dump your model.

10
00:00:45,210 --> 00:00:48,150
So what exactly is a meeting of this company and why?

11
00:00:48,210 --> 00:00:54,450
There is a need, let's say, if you will notice over here, I have simply write a function over here

12
00:00:54,840 --> 00:01:03,960
so that I don't have to do this task for each and every model so that I can automate my own stuff in

13
00:01:03,960 --> 00:01:04,730
a similar way.

14
00:01:04,920 --> 00:01:12,240
If you don't want to create your model again and again and don't want to do all this stuff, just do

15
00:01:12,240 --> 00:01:15,010
dumping and save your model for somewhere else.

16
00:01:15,210 --> 00:01:21,540
So whenever there's a need of that model just loaded using some fancy models of Python, that is exactly

17
00:01:21,540 --> 00:01:25,890
what Bikal modules to just load that model and reuse it again.

18
00:01:26,010 --> 00:01:28,830
So what exactly is inside in this model.

19
00:01:29,100 --> 00:01:32,940
So model exactly contains a mathematical equation model.

20
00:01:32,940 --> 00:01:33,540
Exactly.

21
00:01:33,540 --> 00:01:38,550
Contains all the coefficient with respect to what our algorithm we are going to use.

22
00:01:38,880 --> 00:01:41,390
So let's say if you really want to dump good model.

23
00:01:41,640 --> 00:01:47,130
So here I'm going to say very first, you have to open some file for this.

24
00:01:47,130 --> 00:01:52,290
I'm going to say I'm going to use some while handling techniques already.

25
00:01:52,620 --> 00:01:55,800
So let's say you have to dump your model in some file.

26
00:01:56,040 --> 00:02:00,810
So for this, I have to say I have to open some file in Rightmove.

27
00:02:00,930 --> 00:02:05,580
So very first lesson, I have to open my file, let's say, over here.

28
00:02:05,590 --> 00:02:09,320
So I'm just going to copy the spot from here and here.

29
00:02:09,330 --> 00:02:11,480
Basically, I'm going to pasted.

30
00:02:11,730 --> 00:02:16,400
So after it, we have to mention what can be done.

31
00:02:16,560 --> 00:02:26,730
My model M, so my model is nothing but model dot l so dot pixel is exactly your extension because the

32
00:02:26,730 --> 00:02:32,730
module that will come into picture whenever you have to load your model, whenever you have to dump

33
00:02:32,730 --> 00:02:35,850
your model is exactly your typical model.

34
00:02:35,880 --> 00:02:39,480
So I'm just going to import this pixel model over here.

35
00:02:39,750 --> 00:02:45,570
After what we have to do after doing all these things, what do we have to do and say?

36
00:02:45,570 --> 00:02:51,440
I'm going to say this is exactly my let's say this is my this is exactly my file.

37
00:02:51,540 --> 00:02:57,540
And after this, what I'm going to do, I'm just going to dump using my Bikal model.

38
00:02:57,540 --> 00:03:01,650
And you will see it has a function which is exactly the dump.

39
00:03:01,890 --> 00:03:08,240
And using this function, what I have to do, I have to dump my model in this file.

40
00:03:08,250 --> 00:03:12,660
That's what is the meaning of these parameters that I have.

41
00:03:12,660 --> 00:03:13,540
Bussau, dear.

42
00:03:13,810 --> 00:03:21,000
Let's say I had to mention some condition over here and let's say I don't want to dump model in each

43
00:03:21,000 --> 00:03:21,810
and every case.

44
00:03:21,990 --> 00:03:28,620
So for this, I am just going to add some simple conditionality or just mention it is equally close

45
00:03:28,620 --> 00:03:29,160
to land.

46
00:03:29,670 --> 00:03:32,460
Just dump all these stuff.

47
00:03:32,490 --> 00:03:37,800
Similarly, over here, you have to pass a parameter which is exactly your dump.

48
00:03:37,980 --> 00:03:41,990
And after what we have to do next, we have to play with that.

49
00:03:42,010 --> 00:03:45,660
So we have to play with this for a signature.

50
00:03:45,930 --> 00:03:48,730
Let's it yeah, I have to dump my models.

51
00:03:48,750 --> 00:03:51,280
In such case, I can pass one or that's it.

52
00:03:51,540 --> 00:03:56,910
So if again I'm going to execute it now, you'll see all of this stuff gets executed.

53
00:03:56,910 --> 00:04:04,620
And if you will visit your part, you will see a filename which is exactly what Beaky l file having

54
00:04:04,620 --> 00:04:06,150
will be an extension.

55
00:04:06,420 --> 00:04:10,950
So which had that side and this is exactly the model here.

56
00:04:11,160 --> 00:04:17,620
So this is exactly the model, whatever you have done or whatever you have done already.

57
00:04:17,730 --> 00:04:20,460
So this is exactly the model that you need.

58
00:04:20,610 --> 00:04:27,270
So let's move ahead with our next Pohlman statement in which I have to play with multiple algorithms.

59
00:04:27,280 --> 00:04:33,630
Let's say you can play with this isn't rescan and linear regression and multiple regression algorithm.

60
00:04:33,630 --> 00:04:35,150
That's all approved.

61
00:04:35,180 --> 00:04:41,970
Whatever algorithm you just passed, that algorithm name in your function, that's literally all of

62
00:04:42,090 --> 00:04:42,320
you.

63
00:04:42,510 --> 00:04:46,020
So let's say I'm just going to play with X to the very first.

64
00:04:46,230 --> 00:04:52,020
I'm going to say I have to import some algorithms or let's say I'm going to say the very first algorithm

65
00:04:52,020 --> 00:04:55,800
that I'm going to consider is exactly my, let's say, linear regression.

66
00:04:56,040 --> 00:04:59,910
And after it, I'm going to consider let's say I'm going to play with.

67
00:05:00,490 --> 00:05:09,610
And so I'm going to say from as I could learn, I have to import these neighbors and here I have a class

68
00:05:09,610 --> 00:05:16,720
which is my key neighbors, because it is a regression, you suggest, and after it, I am also going

69
00:05:16,720 --> 00:05:17,800
to play with dissidents.

70
00:05:17,960 --> 00:05:27,820
So I am going to say from this cycle on, I'm going to say, gee, I have to import my decision, tree

71
00:05:27,880 --> 00:05:28,570
regulation.

72
00:05:28,900 --> 00:05:36,550
So just like you ordered and in this Graddick function, you have to just pass this linear regression

73
00:05:36,550 --> 00:05:38,630
and you have to initialize it as well.

74
00:05:38,950 --> 00:05:40,510
So just execute it.

75
00:05:40,510 --> 00:05:43,720
You will see or hear all of this stuff gets executed.

76
00:05:43,720 --> 00:05:47,490
And this is exactly the situation in case of linear regression.

77
00:05:47,500 --> 00:05:52,890
And if you are thinking why I have zero because I don't have to dump my model.

78
00:05:53,080 --> 00:06:00,040
Similarly, you can play with the word like say I'm going to try and cry with my left decision.

79
00:06:00,690 --> 00:06:03,060
I have to just call this function here.

80
00:06:03,070 --> 00:06:05,130
I have to pass my decision tree.

81
00:06:05,140 --> 00:06:05,620
Let's see.

82
00:06:06,340 --> 00:06:09,420
I have to import my decision tree redressal.

83
00:06:09,430 --> 00:06:17,180
So here I'm going to say this integer agressor here, I have to simply initialize my decision tree requested.

84
00:06:17,200 --> 00:06:19,540
So you have to execute it as well.

85
00:06:19,780 --> 00:06:23,620
And here you are just going to initialize this regression.

86
00:06:23,830 --> 00:06:26,010
And I don't have my model.

87
00:06:26,020 --> 00:06:26,780
It's all up to you.

88
00:06:26,800 --> 00:06:28,060
So just execute it.

89
00:06:28,330 --> 00:06:33,270
It will take a long time and you will see it had that much accuracy in case of decision tree.

90
00:06:33,610 --> 00:06:39,700
It has that much autist, good gurpreet and all these different kind of matics in case of decision.

91
00:06:39,850 --> 00:06:44,860
And all these are my predictions with respect to why this isn't the algorithm.

92
00:06:44,860 --> 00:06:49,060
And this is exactly a beautiful distribution in case of decision.

93
00:06:49,270 --> 00:06:55,270
And this is the distribution in case of linear regression that has somewhere around sixty three percent

94
00:06:55,270 --> 00:06:55,840
accuracy.

95
00:06:55,840 --> 00:07:02,650
So it's still you will notice my random forest because it has approx 80 percent accuracy.

96
00:07:02,680 --> 00:07:06,260
Similarly, you can play with multiple algorithm.

97
00:07:06,280 --> 00:07:07,750
That's all up to you.

98
00:07:07,760 --> 00:07:10,570
Let's say I'm going to play in the station.

99
00:07:10,870 --> 00:07:13,270
So nearest neighbor.

100
00:07:13,270 --> 00:07:16,210
So I'm going to say Kinabalu regression.

101
00:07:16,360 --> 00:07:19,210
I don't have to dump my model just by zero.

102
00:07:19,480 --> 00:07:26,140
Just if you order, it will take up a lot of time and you will see it has somewhere around sixty six

103
00:07:26,140 --> 00:07:27,250
percent accuracy.

104
00:07:27,460 --> 00:07:35,800
So whenever your data set is huge, never go with Kinen because again, it is not advisable to get a

105
00:07:35,800 --> 00:07:40,120
scientist whenever they are working on large data sets.

106
00:07:40,120 --> 00:07:47,230
Similarly, you can play with your SVR and multiple regression algorithms that are available there.

107
00:07:47,680 --> 00:07:49,160
So that's all about this session.

108
00:07:49,290 --> 00:07:50,440
Hopefully will have the session.

109
00:07:50,440 --> 00:07:51,520
Very much so.

110
00:07:51,520 --> 00:07:52,180
Thank you.

111
00:07:52,180 --> 00:07:53,160
Have a nice day.

112
00:07:53,170 --> 00:07:54,190
Keep learning.

113
00:07:54,430 --> 00:07:55,390
Keep going.

114
00:07:55,660 --> 00:07:56,620
Keep practicing.