Classification methods in machine learning
There is a certain classification task: for training, the classifier receives an array of strings as a class and some numbers as parameters.
Are there any methods, without using neural networks, that allow you to classify such samples? Which direction to go?
Example:
Training:
RW, LW, CAM -> 75,45,22,78,98,76
CB -> 56,54,66,77,25,56
CB,RB,RCB -> 45,56,78,97,44,22
CAM,CM,CDM -> 34,45,65,77,88,93
Prediction:
34,56,76,88,99,23 -> CAM,CM,RW
24,56,90,98,77,66 -> ST,LW,RW
34,45,87,98,77,55 -> RB,CB
I can't solve the problem, because I don't know about methods with multiple classes for an object yet.
I will be grateful for your advice!
1 answers
This is a completely classical, one might say - the basic problem of classification theory. It is described in any not even a book, but in any blog on machine learning. Be in one click: https://ru.wikipedia.org/wiki/Задача_классификации https://tproger.ru/translations/scikit-learn-in-python/ https://evergreens.com.ua/ru/articles/classical-machine-learning.html https://wiki.loginom.ru/articles/classification-problem.html http://www.aiportal.ru/articles/autoclassification/methods-class.html
, etc. But which method to use is decided on the basis of both the corresponding theoretical constraints and the comparison of the results of their application.