Matlab Optimi Toolbox

This page describes the optimi toolbox for MATLAB available for download here.

Dependencies

The optimi toolbox depends on the ndlutil toolbox. You can use your user name and password for accessing these toolboxes.

Release Information

Current release is 0.132.

Included a function to return the default functions to be used for particular constraints and a function to return the default optimiser.

Added wrappers for using Carl Rasmussen's conjugate gradient code.

This toolbox allows computation of several different non-linearities and associated conversions and their gradients.

There are several kernels implemented, the ones that are being maintained for the latest release are:

exp sigmoid negLogLogit

exp constrains a parameter to be positive through exponentiatiation. sigmoid constrains a parameter to be between 0 and 1 through the sigmoid function and negLogLogit constrains a parameter to be positive through the function log(1+exp(x)). In much of the code based on this we prefer negLogLogit to exp as it is better behaved for large x..