Welcome to PyBrain’s documentation!

The documentation is build up in the following parts: first, there is the quickstart tutorial which aims at getting you started with PyBrain as quickly as possible. This is the right place for you if you just want get a feel for the library or if you never used PyBrain before.

Although the quickstart uses supervised learning with neural networks as an example, this does not mean that that’s it. PyBrain is not only about supervised learning and neural networks.

While the quickstart should be read sequentially, the tutorial chapters can mostly be read independently of each other.

In case this does not suffice, we also have an API reference, the Module Index. Most of the packages and modules of PyBrain are auto-documented there.

If you want to develop for PyBrain and contribute, check out our guidelines in our wiki: http://wiki.github.com/pybrain/pybrain/guidelines.

If at any point the documentation does not suffice, you can always get help at the pybrain Google Group at http://groups.google.com/group/pybrain.

Installation

Quick answer:

$ git clone git://github.com/pybrain/pybrain.git
$ python setup.py install

Long answer: We keep more detailed installation instructions (including dependencies) up-to-date in a wiki at http://wiki.github.com/pybrain/pybrain/installation.

Quickstart

• Building a Network
• Building a DataSet
• Training your Network on your Dataset

Tutorials

• Introduction
• Building Networks with Modules and Connections
  • Feed Forward Networks
  • Examining a Network
  • Naming your Networks structure
  • Using Recurrent Networks
• Classification with Feed-Forward Neural Networks
• Using Datasets
  • superviseddataset
  • sequentialdataset
  • classificationdataset
  • importancedataset
• Black-box Optimization
  • Continuous optimization
  • General optimization: using Evolvable
  • Optimization in Reinforcement Learning
• Reinforcement Learning
• Extending PyBrain’s structure
  • Layers
  • Connections
  • ParameterContainers
  • Checking for correctness

Advanced

• Fast networks for PyBrain
  • Installation
  • Usage
  • Limitations
• Using ODE Environments
  • Using an existing ODE environment
  • Creating your own learning task in an existing ODE environment
  • Creating your own ODE environment
  • Creating your own XODE instance

API

• connections – Structural Components: Connections
• evolvable – Container Component: Evolvable
• modules – Structural Components: Modules
• networks – Structural Components: Networks
• actionvalues – RL Components: ActionValues
• agents – RL Components: Agents
• experiments – RL Components: Experiments
• explorers – RL Components: Explorers
• learners – RL Components: Learners
• tasks – RL Components: Tasks
• optimization – Black-box Optimization Algorithms
• classification – Datasets for Supervised Classification Training
• importance – Datasets for Weighted Supervised Training
• sequential – Dataset for Supervised Sequences Regression Training
• supervised – Dataset for Supervised Regression Training
• svmunit – LIBSVM Support Vector Machine Unit
• svmtrainer – LIBSVM Support Vector Machine Trainer
• trainers – Supervised Training for Networks and other Modules
• tools – Some Useful Tools and Macros
• utilities – Simple but useful Utility Functions
• nearoptimal – Near Optimal Locality Sensitive Hashing

Indices and tables

• Index
• Module Index
• Search Page