Modular toolkit for Data Processing MDP

Abstraction: Python modular ML pipeline framework with PCA, ICA, and manifold methods

Key points:

• MDP is a Python framework for building supervised/unsupervised learning pipelines and feed-forward network architectures from reusable processing units
• Algorithms include: PCA, ICA, Slow Feature Analysis, Hessian Locally Linear Embedding, Factor Analysis, Restricted Boltzmann Machines (RBM), and various classifiers
• New algorithms are easy to implement and auto-integrate with the rest of the library
• BSD licensed; targets scientific developers, educators, and researchers
• Hosted on SourceForge; Python-only

Connections: Mdp Toolkit · Machine Learning · Data Processing

Source: http://sourceforge.net/projects/mdp-toolkit/?source=dlp