Welcome to sknn’s documentation!

Deep neural network implementation without the learning cliff! This library implements multi-layer perceptrons as a wrapper for the powerful pylearn2 library that’s compatible with scikit-learn for a more user-friendly and Pythonic interface.

---

Module Reference

• sknn.mlp — Multi-Layer Perceptrons
  • Layer Specifications
  • MultiLayerPerceptron
  • Regressor
  • Classifier

User Guide

• Installation
  • A) Download Latest Release [Recommended]
  • B) Pulling Repositories [Optional]
• Model Setup & Training
  • Regression
  • Classification
  • Convolution
  • Per-Sample Weighting
  • Native & Custom Layers
• Low-Level Configuration
  • Keyboard Interrupt
  • CPU vs. GPU Platform
  • Multiple Threads
  • Backend Configuration
• Input / Output
  • Verbose Mode
  • Saving & Loading
  • Extracting Parameters
• scikit-learn Features
  • sklearn Pipeline
  • Grid Search
  • Randomized Search
  • Unsupervised Pre-Training
• Customizing Learning
  • Training Callbacks
  • Epoch Statistics

Indices & Search

• Index
• Search Page

---