hdnet v0.1 manual

hdnet stands for Hopfield denoising network. It is a Python package for analysis of neural population spiking data, i.e. parallel spike trains.

In particular, it provides a novel method for finding and extracting salient low-dimensional representations of the dynamics of populations of spiking neurons based on a denoising approach to spatiotemporal patterns (STP) contained in the data.

An always up-to-date online-version of the documentation can be obtained at

http://team-hdnet.github.io/hdnet/

This manual also can be downloaded in a PDF version

Contents:

• Introduction
• Installing hdnet
  • Adding OpenBLAS support
• Citing hdnet
• Mathematical background
  • Hopfield Networks
  • Dichotomized Gaussian
• Basic example
  • Starting off
  • Going further
  • Saving and loading
  • Stimuli
  • Real data
• Example: Analyzing pattern sequences
  • Creating a synthetic data set
  • Fitting a Hopfield network
  • Examining the pattern sequence
  • Constructing the Markov graph
  • Indentifying base states
  • Cycles as reliably produced network reponses
• hdnet package
  • hdnet.data
  • hdnet.hopfield
  • hdnet.learner
  • hdnet.math
  • hdnet.patterns
  • hdnet.sampling
  • hdnet.spikes
  • hdnet.spikes_model
  • hdnet.stats
  • hdnet.stimulus
  • hdnet.util
  • hdnet.visualization
• References

Indices and tables

• References
• Index
• Module Index
• Search Page