Code and results for the following problems:
-
Implementing a custom version of Value Iteration and computing the Optimal Policy for a Markov Decision Process.
-
Creating a decision tree based on attributes used to determine acceptance of students into graduate school.
-
Building a support vector machine that uses two-class training data, and extracting parameters from it.
-
Replicating the learning process of finding a linear separator by starting with a random perceptron and updating.
-
Determining the minimum error for a single perceptron, and developing a system of connected, multi-layered perceptrons.