Binary Hyper-parameter Decision

October 22, 2015

Report Abuse
Binary classification tasks require optimal parameterization using a training data set. In this module, we automate the process of parameterization across 4 well-known binary classifiers (k-nearest neighbors, support vector machines, decision trees and logistic regression) to maximize the classification accuracy on a separate test dataset.
[Click for a detail description of the binary hyper-parameter decision flow][1] [1]: