The Machine Learning module aims to implement just about all popular techniques and methods of predictive related algorithms a little bit at a time. At the moment several functions in their respective files are stubbed out within the ml module header directory.
Status:
Linear Regression: Semi-Working on two column datasets. Algorithm is promising but selection of dependent/independent variables is needed. This is where the proposed #29 DataTable functionality could come into play this should be an option for all functions in openMTPK.
Multiple Linear Regression: This could possibly be spawned off from the Linear Regression implementation in a template like approach.
Activations, Trainers, Regularizers, and Optimizers are all housed in their respective files. These files aim to implement relevant functions for working across a range of ML related functions.
Neural Networks: Self-Organizing Maps (Kohonen), Multi-Layer Perceptron, Bayes Network, and more are goals of the Machine Learning Module.
In all, think of this module of openMTPK as a low level implementation of Scikit-learn. Sklearn features modular prebuilt methods for a slew of Machine Learning algorithms.
The Machine Learning module aims to implement just about all popular techniques and methods of predictive related algorithms a little bit at a time. At the moment several functions in their respective files are stubbed out within the ml module header directory.
Status:
In all, think of this module of openMTPK as a low level implementation of Scikit-learn. Sklearn features modular prebuilt methods for a slew of Machine Learning algorithms.