mrscake is a machine learning library that automatically picks the best model for you.

mrscake is a machine learning library for building classifiers from labelled training data. It has support for over a dozen different machine learning models, and also supports automatic model selection as well as optimized code generation for all models.

Most of the machine learning models are adapted from OpenCVs machine learning library.

Supported models

The current version has support for linear classifiers, perceptrons, multilayer perceptrons (artificial neuronal networks) with linear, gaussian and sigmoid activation functions, support vector machines with linear, sigmoid and rbf kernels, k-nearest neighbors, decision trees and decision tree forests (random trees, extremely randomized trees, gradient boosted trees.)


The easiest way to get started is by reading the tutorial:
mrscake tutorial

mrscake has interfaces in C, Python and Ruby:
Python interface documentation
Ruby interface documentation
C interface (header file)