Books classifier
At first as a business number, the second as a business book provided by different combination types (classes), the third column is the total price after combination,
O how according to which the merchants of randomly generated categories of books classification
Such as
A class of merchants have,, the price is for,,
The merchants of class B,, the price is,,
Businesses have a class AB,, the price is for,,
AD class merchants have,, the price is for,,
. ...
. ...
EACBD class merchants have? , the price? ,
Note: AB and BA is a kind of, together
ABC, BCA, CBA is a kind of together
There seems to be a variety of combination way
Like
Use matrix
Want to output all the classification results of
Links below for the project, which is randomly generated book stores information


The picture is not done
Please help agghh
Integration is not only 20,

CodePudding user response:

https://pan.baidu.com/s/1B_O5mNEVboOLfGqmQCmYmQ

CodePudding user response:

reference 1/f, small _ @ reply:

https://pan.baidu.com/s/1B_O5mNEVboOLfGqmQCmYmQ

This is the project link, ask everybody to help

CodePudding user response:

Descartes sets again?? Oh, it's not already combination, need to determine "feature similarity", using "cosine similarity" (ignore the distance, only determine Angle)

CodePudding user response:

https://blog.csdn.net/weixin_43666029/article/details/89306480

CodePudding user response:

Not, just put the generated as a result, into the matrix, and then in the judgment

CodePudding user response:

reference wanghui0380 reply: 3/f

Descartes sets again?? Oh, not have combination, needs to determine "feature similarity", using "cosine similarity" (ignore the distance, only determine Angle)

Not, just put the generated as a result, into the matrix, and then in the judgment, with similarity is the class

CodePudding user response:

B. a. string sorting. Indexof (string) actually so, anyway you want to do is to ignore the distance, the practice of sorting is normalized, the character encoding sequence when the vector, one directly, so distance from nothing,

CodePudding user response:

refer to 6th floor small _ @ reply:

Quote: refer to the third floor wanghui0380 response:

Descartes sets again?? Oh, not have combination, needs to determine "feature similarity", using "cosine similarity" (ignore the distance, only determine Angle)

Not, just put the generated as a result, into the matrix, and then in the judgment, with similarity is class ah

Similarity is 1, ABC and BAC under the condition of the distance to be ignored, they are same, similarity of 1

CodePudding user response:

Using similarity is the class, saying, now called artificial intelligence, is exactly what do similarity classifier, whether

See git a star on the artificial intelligence learning strategy

Supervised learning
Data preprocessing
Simple linear regression
Multiple linear regression
Logistic regression
K neighbor method (k - NN)
Support vector machine (SVM)
The decision tree
Random forest
Unsupervised learning
K - average clustering
Hierarchical clustering

CodePudding user response:

refer to the eighth floor wanghui0380 response:

Quote: refer to the sixth floor small _ @ reply:

Quote: refer to the third floor wanghui0380 response:

Descartes sets again?? Oh, not have combination, needs to determine "feature similarity", using "cosine similarity" (ignore the distance, only determine Angle)

Not, just put the generated as a result, into the matrix, and then in the judgment, with similarity is class ah

Similarity is 1, ABC and BAC under the condition of the distance to be ignored, they, similarity of 1

Ok, thank you, I have widened the field of vision, have never heard of this before

CodePudding user response:

Descartes method can solve a look at this is you write UI to written in asynchronous processing such data breakpoint debugging line don't show the UI will be very slow if the amount of data is very large.

CodePudding user response:

11 references Bycnboy response:

cartesian method can solve a look at this is you write UI to written in asynchronous processing such data breakpoint debugging line don't show the UI will be very slow if the amount of data is very large.

But still not ah -_ - | |, can help you to write down a

CodePudding user response:

11 references Bycnboy response:

cartesian method can solve a look at this is you write UI to written in asynchronous processing such data breakpoint debugging line don't show the UI will be very slow if the amount of data is very large.

Flute, card, no, no, no, I mean you understand wrong? Is that these letters is to keep those generated to classification, and not let their group, or whether it's a letter, or more letters, is to make their classification, classified according to a letter the same letter, two would be classified by two identical letters, three words as three of the same, as such, and is mainly put the letters into those for randomized, classification for him, rather than put it, to combine like taobao shopping to buy clothes, well, the size code color combination,

Now the problem to be solved is how to classify him? Or is it every endures a 5 by 5 matrix inside, well, if only a, is all of the back four is empty or other values, if B in front of the first is empty, the third behind is empty or other values, if have, is the row, is abcde, this method is on the inside the matrix, to solve the problem is the problem

CodePudding user response:

With k - mean clustering, sklearn under the python library, ready-made