Articles’ software code
Below you will find a list of available open-source libraries, organized by topic, that you can use to repeat several experiments taken from my research papers, or as a starting point for further explorations. Some of them require the installation of the Lynx toolbox. Most of them are implemented in MATLAB, some in Python 2.7/3.5.
- Distributed RVFL networks with data distribution and model distribution (both require Lynx).
- Distributed Echo state networks based on ADMM. Part of this code was adapted from an ESN toolbox by H. Jaeger.
- Distributed semi-supervised algorithms for linear support vector machines and kernel ridge regression via matrix completion.
- A distributed version of the spline adaptive filter for nonlinear estimation. If you are interested in spline filters, you can also check the recently released SAF Toolbox from our research group.
- Distributed algorithms for general neural networks (under additional development, contains a full Python porting with Theano).
- A Python library for group sparse regularization (allowing to remove entire neurons during the optimization phase).
- MATLAB library for adapting activation functions based on cubic spline interpolation.
- Online Sequential ELM with Kernels (require Lynx). If you are working with ELM, please consider the recent controversy surrounding its origins.
- Semi-supervised RVFL Network (require Lynx).