Hello Nabila,

First of all, the data set has too much less data points. Hardly any good classifier may work well on this. So, my suggestion is to increase more observations for better classification accuracy.

Thanks,

Sobhan

Hello Nabila,

First of all, the data set has too much less data points. Hardly any good classifier may work well on this. So, my suggestion is to increase more observations for better classification accuracy.

Thanks,

Sobhan

Hi Nabila,

There is no one algorithm that works well all the time.

If your player type is nominal, then this is a classification problem. Several algorithms can be used including the following example algorithms:

-Naive Bayes

-LibLINEAR

-LibSVM

-Multinomial logistic regression

-SVM

-K-nearest neighbours

-Decision tree

If your player type is numeric (regression problem), you could tray the following example algorithms:

-Gaussian Processes

-Linear Regression

-Support vector machine for regression

-M5 Model trees

-Multilayer Perceptron

Compare several algorithms and select the one produces the best result.

NB: Evaluate each learning process using, e.g., cross-validation. Try to increase the learning set. Remember that there are many other tools to try (not only those that I listed up). You need to do some reading about your topic in general.

HTH.

Samer Sarsam, PhD.

Dear Nabila Hamdaoui,

Look the link, maybe heilful.

http://www.jmlr.org/papers/volume10/chen09a/chen09a.pdf

https://digitalcommons.unl.edu/cgi/viewcontent.cgi?referer=https://www.google.az/&httpsredir=1&article=1086&context=computerscidiss

https://arxiv.org/pdf/1604.08075.pdf

Regards, Shafagat

Hello

You can try fuzzy or anfis models.

Regards, Pascual.