Simple machine learning algorithms implemented in python during a class at Winchester High School.
Each directory is a separate project.
Artificial neural nets -
ann - a form of function approximation learning.
The project includes a neural net for optical character recognition.
Genetic algorithms -
ga - optimize a given function. The hungarian
ha - is a specific algorithm to maximally select elements of a
matrix. We used it in combination with a genetic algorithm to create
a program to optimize math team performance.
Hidden markov models -
hmm - perform pattern recognition. From an extracted
model, the Viterbi algorithm allows prediction of underlying states from
Decision trees -
tree - classify and predict from learned data. We applied
the decision tree program to car buyers, the census, congressional voting
records, and CPU performance data.