Which would be a suitable model when an application has to classify people as native english language speakers or not based on their signatures?

Note: All signatures are in english language.

Regression model

Clustering model

Classification model