.

Contributed Software

This page contains links for downloading files that have been contributed by other developers. Eventually, these files may be integrated with the main Netlab toolbox - but that takes time, so the files have been made available here in the interim:
  • Windowing in hybrid Monte Carlo sampling derived from the C code in Radford Neal's FBM software (with his permission). Software written by Aki Vehtari.
  • Efficient training of RBF classifiers using logistic/softmax output and IRLS training algorithm. Sample code used to generate performance results in IJNS paper. Software written by Ian Nabney.
  • Visualisation tool using hierarchical GTM. Software written by David Evans, Ian Nabney, Yi Sun and Peter Tino.
  • Gaussian Mixture Model using log probabilities (so as to avoid numerical problems in high-dimensional datasets). Software written by Alexander G Dimitrov.
  • Parallel tempering for sampling multi-modal distributions.  Software and four example scripts.  Also includes optBINS Matlab code by K. Knuth.  Software contributed by Laurent de Vito.

Employable Graduates; Exploitable Research