Classification problem

machine_learning
predictive_model
data_science

#1

I’ve 50 variables and 28 are numerical and 16 are categorical and 6 are binary (1/0), what is best algorithm to predict Yes/No ( classification).

I tried with logistic regression but none of the categorical variables are significant in the model output.

Thanks


#2

hi @sagar4dk

Maybe you can try using random forest model, or either go with some dimensionality reduction technique first and then apply any machine learning algorithm.

You can also check out Catboost algorithm which handles categorical variables automatically.

Here is the link for the catboost article. https://www.analyticsvidhya.com/blog/2017/08/catboost-automated-categorical-data/

Cheers!
Shubham