Release Notes: This release was focused on fixing bugs and improving usability.
Release Notes: This release adds relevance vector machines for solving regression and classification problems.
Release Notes: The support vector classification and training tools have been changed to use a simpler and more powerful interface. This release also adds tools for training radial basis function networks, performing feature ranking, and approximating an SVM decision function with a reduced set of support vectors.
Release Notes: This release was focused on improving existing library components. A few bugs were fixed. The Unicode support in the GUI toolkit was improved. And some of the machine learning algorithms were made easier to use.
Release Notes: This release adds two new kernel learning algorithms. The first is a centroid estimator/novelty detector and the second is a non-linear k-means clustering algorithm. This release also adds some unconstrained optimization routines to the library, so it now contains a conjugate gradient optimization algorithm as well as a quasi-newton algorithm.
Release Notes: This release adds a new kernel based regression algorithm to the set of machine learning algorithms already present in the library. It also includes some minor bugfixes.
Release Notes: This release includes minor improvements to the GUI widgets and matrix utilities. It also includes a graphical application for drawing and performing inference in Bayesian Networks.
Release Notes: This release adds two new GUI widgets, the text_grid and scrollable_region objects. It also fixes some bugs in the GUI code as well as a bug in one of the Bayesian Network inference algorithms.
Release Notes: This release was mainly focused on fixing some minor bugs in the GUI widgets.
Release Notes: This release of the library includes significant improvements to its GUI features. In particular, there is now menu bar and popup menu support as well as user-selectable button styles and tooltips.