This project contains a neural network all built within Java. It is tested on a 2D dataset with x and y variables. Each variable is between 0-50, if the x variable is under 33 and the y variable is under 23 then a datapoint is safe. The goal of the model is to predict whether or not each datapoint is safe. With the given parameters and hyperparameters the model takes approximately 100,000 epochs to reach an accuracy of ~100%. I expect to continue to add more functionality to the project, adding different activation, cost, and normalization functions. I also expect to add functionality for different datasets, such as the MNIST dataset.
-
Notifications
You must be signed in to change notification settings - Fork 0
ReilleyMilne/NeuralNetwork
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
Created a basic neural network within Java. Made custom classes for activation, cost, and normalization functions. Includes classes for processing data and saving models.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published