We study the flavor dependence of the Chiral Magnetic Effect (CME) using two of the primary correlators used to characterize the charge separation effect. These are the correlator \(\varDelta \gamma { andthecorrelator}R_{\psi _{2}} \) . We use the AMPT (A Multiphase Transport Model) model to study the sensitivity of these correlators to two and three flavors of quarks. The AMPT model used has a centrality dependent charge separation introduced in the initial stage. We find that the \(R_{\psi _{2}} { correlatorshowsbetterflavordependencythanthetraditional}\varDelta \gamma \) corrrelator. We then create a classification model with a neural network architecture and train the model using numerous combinations of the final state particle distributions.