machine learning - Can a SVM be used to differentiate samples ? -

I am working for the problem, I have to know whether two samples are identical.

I want a class 1 when two samples are equal and class 0, when they are not.

I do not know how it has been explained better, so this is a stupid example:

Training set: {A, B}, {C, D}, e results : AB: 1, AC: 0, AD: 0, AE: 0, BA: 1, BC: 0, BD: 0, BE: 0, ...

I have only two set features Svm is added, but the problem is that memory is necessary (n²) which is very large: (

Is there any way to solve this type of problem with SVM?

PS I know that I I can use the label, it is not what I am looking for.

classification of discrimination What SVM does is. If you do not want to label your data, then you will not be able to apply the methods of supervised learning, however depending on your specific problem you will select an unstable learning algorithm. / P >

However, make sure that you understand that What is the meaning of 'belling'? When you say

Result: AB: 1, AC: 0, AD: 0, AE: 0, BA: 1, BC: 0, BD: 0, BE: 0, ...

then give your data (binary) label if you define what the similarities are and provide training data If you are able, you can learn to be supervised.

Also, keep in mind that a space complexity of Ω (n²) is not something that you are able to beat with most machine learning techniques because high number of approaches to data and / or learning To use the matrix. The experience has also shown that it is not unrealistic.