How can I justify the a priori choice of a classifier compared to another one?

How can I justify the a priori choice of a classifier compared to another one? Which metrics or experiments can I use as justification for using one classifier rather than another one?

I’m dealing with numerical data and I am a beginner with data mining.

Thanks in advance
Antonio

Could you see if this entry also can help you with your question? Or you might tell us in more details what you want to do.

1 Like

thanks for your reply, it was useful. I try to explain me better: I built a three-class classification model that works good with XGB Tree Learner because I got over 95% in accuracy . Neverthless during the presentation of my work I will have to illustrate the reasons why I chose this model instead of others. I need a comparative analysis using the same metrics and prove that my model was the best one. I hope I was clear now.

Coud you help me?

I read in my DM book that I could use the ten-times ten-fold cross validation. But how?

Thanks in advance
Antonio