abagail.html

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <title>ABAGAIL</title>
</head>


<body bgcolor="#D4D9C3">

<font face="Arial,Helvetica">

<a href="index.html">Go back to Andrew Guillory's home page</a>

<h1>ABAGAIL</h1>

<b>the <i>Absolute Best</i> Andrew Guillory Artificial Intelligence Library</b>

<br>
<p>This library is the result of close to a year of research
and class work in artificial intelligence (AI).  It contains
a number of interconnected Java packages that implement
machine learning and artificial intelligence algorithms.
These are artificial intelligence
algorithms implemented for the kind of people that
like to implement algorithms themselves.</p>

<b>June 27, 2005</b><p>Fixed a bug in the SVM code, expanded
the reinforcement learning package.</p>

<a href="ABAGAIL.zip">Download Java Source Code</a>
<br/>
<h2>Features</h2>

<h3>Hidden Markov Models</h3>
<ul>
<li>Baum-Welch reestimation algorithm, scaled forward-backward
algorithm, Viterbi algorithm</li>
<li>Support for Input-Output Hidden Markov Models</li>
<li>Write your own output or transition probability distribution
or use the provided distributions, including neural network
based conditional probability distributions</li>
</ul>

<h3>Neural Networks</h3>
<ul>
<li>Feed-forward backpropagation neural networks of 
arbitrary topology</li>
<li>Configurable error functions with sum of squares, 
weighted sum of squares</li>
<li>Multiple activation functions with logistic sigmoid, 
linear, tanh, and soft max</li>
<li>Choose your weight update rule with standard update rule, 
standard update rule with momentum, Quickprop, RPROP</li>
<li>Online and batch training</li>
</ul>

<h3>Support Vector Machines</h3>
<ul>
<li>Fast training with the sequential minimal optimization 
algorithm</li>
<li>Support for linear, polynomial, tanh, radial basis 
function kernels</li>
</ul>

<h3>Decision Trees</h3>
<ul>
<li>Information gain or GINI index split criteria</li>
<li>Binary or all attribute value splitting</li>
<li>Chi-square signifigance test pruning with configurable 
confidence levels</li>
<li>Boosted decision stumps with AdaBoost</li>
</ul>

<h3>K Nearest Neighbors</h3>
<ul>
<li>Fast kd-tree implementation for instance based algorithms of 
all kinds</li>
<li>KNN Classifier with weighted or non-weighted classification, 
customizable distance function</li>
</ul>

<h3>Linear Algebra Algorithms</h3>
<ul>
<li>Basic matrix and vector math, 
a variety of matrix decompositions based on
the standard algorithms</li>
<li>Solve square systems, upper triangular systems, 
lower triangular systems, least squares</li>
<li>Singular Value Decomposition, QR Decomposition, 
LU Decomposition, Schur Decomposition, Symmetric Eigenvalue Decomposition,
Cholesky Factorization</li>
<li>Make your own matrix decomposition with the easy to use
Householder Reflection and Givens Rotation classes</li>
</ul>

<h3>Optimization Algorithms</h3>
<ul>
<li>Randomized hill climbing, simulated annealing,
genetic algorithms, and discrete dependency tree MIMIC</li>
<li>Make your own crossover functions, mutation functions, neighbor
functions, probability distributions, or use the provided ones.</li>
<li>Optimize the weights of neural networks and solve
travelling salesman problems</li>
</ul>

<h3>Graph Algorithms</h3>
<ul>
<li>Kruskals MST and DFS</li>
</ul>
<h3>Clustering Algorithms</h3>
<ul>
<li>EM with gaussian mixtures, K-means</li>
</ul>

<h3>Data Preprocessing</h3>
<ul>
<li>PCA, ICA, LDA, Randomized Projections</li>
<li>Convert from continuous to discrete, discrete to binary</li>
</ul>

<h3>Reinforcement Learning</h3>
<ul>
<li>Value and policy iteration for Markov decision processes</li>
</ul>


</font>
</body>
</html>