Let’s understanding the way of making ML like making a food

 

What I want to introduce is the way of making model that was called Machine Learning(ML).

There are many site to learn how to use ML. For programmer that process is easy to understand. However, man who do not know those system is difficult to understand. Therefore, in this time I want to explain this by a series of flows like making foods.

In some sites, they explain the definition of ML is not clear for begineer. I think it is so confused for you.

Machine learning is an application of artificial intelligence (AI) that provides systems the ability to automatically learn and improve from data without being explicitly programmed. Machine learning focuses on the development of computer programs that can access data and use it to learn by themselves.

There are many example that ML was used in our dairy life so, before proceed our main story, I want you search in google ‘ML example’ . If you available.

Index

1, collecting some foods you need      1, collecting some data you need

2, cutting those foods to cook  2, preprocessing data 

3, adjusting fire to cook         3,  tuning learning rate

4, tasting and evaluating the taste     4, looking some graphs to evaluate the result

5,conclusion

In here, I want to use unique format to divide situation. In the below, First, I will write situation in cooking and in the next index, I want to introduce ML steps. Please care about that. 

1, collecting some foods you need      1, collecting some data you need

0, deciding what you want to eat   0, deciding what do you want to predict Today, we want to make curry. First we have to collect vegetable and beef.    1, collecting some foods you need    First we have to collect vegetable and beef. 

We want to predict the prices like house or other somethings. 1, collecting some data you need. In the ML , we have to collect house’s data like size, money and the place.(In detail,  we can use SQL that is programming of language focus on collecting data )

2, cutting those foods to cook  2, preprocessing data 

Next is cutting those vegetable that make you easier to cook when you use fire or eat foods. we want to cut those be good size that we can eat. 

Data is also have to be arranged by cutting some bad dates.  For example those data are biased then we have arrange data.(There are NA in the data  so we have to put out this number and fill some good number.)

Id

YearBuilt

LotArea

LotFrontage

1

2003

8450

65

2

1976

9600

80

3

2001

11250

68

4

1915

9550

60

5

2000

14260

84

6

1993

14115

85

7

2004

10084

75

8

1973

10382

NA

9

1931

6120

51

10

1939

7420

50

11

1965

11200

70

12

2005

11924

85

13

1962

12968

NA

14

2006

10652

91

15

1960

10920

NA

 

3, adjusting fire to cook         3,  tuning learning rate

Next is using spices to decide the taste of curry. You know, we do not know what material is in the spices but we know this make curry better taste.  In the ML is also happened the thing. After firing, the curry will be yummy taste.

ML also have spices to improve the taste that is a neural network. we just only have to input clear data. After training This tools will output good prediction.(we can use libraries like pytorch or keras. Those tools is like good pot in cooking. If you use high price tools to cook then your curry will be better than usual. ) Tuning Learning rate is important when we use neural network. It’s like adjusting fire when you cook. In cooking we have to adjust good fire for foods not be burned. If you fail that process your bread will like be below 

4, tasting and evaluating the taste     4, looking some graphs to evaluate the result

When you finish making curry, you have to taste that is good or not. This step is very important for cooker. They deserve good taste.

ML also have the step to check the taste. We have to check the output is  overlapping or not and other many things. There are many factor that improve the model or not , so we have to quote those fact from graph and some output.

5,conclusion

In the final, I want to show you some drawing that I draw up. Those drawing maybe will help you to understand ML process clearly.

If you are interested in ML, I recommend you check kaggle. This site is very useful to learn ML and it have many materials to quick learn.

reference :  kaggle.com