Developed a multi-layer Artificial Neural Network using backpropagation with the help of nonlinear sigmoid functions and stochastic gradient descent rule. With it, we will "train" the program to be able to identify if a mushroom is poisonous or edible with some consistency, based off its characteristics. Coded in Java
To run the program, all you need is the “ANN.java” file and the “mushroom.data” file. Use the “mushroom.data” file as the command line argument and watch the results come in. Built in jGrasp. Ran in both jGrasp and terminal via school computers running Ubuntu Linux.
See the discussion file to read more about how the program works!