Bayesian Applet This applet is
very attractive as it allows us to see the classes and how decision surface will effect with
the change in their variance.
Change X and Y standard deviations of class 1 and see how decision surface changes.
Change X and Y standard deviations of class 2 and see how decision surface changes.
Change X and Y standard deviations of class 1&2 such that they overlap and see how decision surface changes.
Elastic Net for TSP
SOM and elastic nets can be regarded as competitive learning with a topological constraint.
TSP is the most notorious one in the NP-complete problems.
The definition is simple: Find the shortest closed-path through all points.
(You can also see it here.)
Bayesian Self-Organizing Maps (BSOM)
SOM as a statistical model. Learning is regarded as an estimation algorithm for its parameters.
Hyperparameters also are estimated. Ultimately a probability densty function for data is estimated.
Test of Unicursality of a Graph
(in Japanese)
First, set vertices of a graph. Then, press the next button and go to the link setting mode.
Finally, by pressing the next button the possibility to draw the graph with a single stroke of
the blush is determined. The author has other applets.