The project is build to perform handwritten digit recognization using machine learning preceptron algorithm. The
model is developed using Java OOP concepts. The test and train data can be used to perform the respective
functions. The model developed works with an accuracy of 99% approximately.
The project is developed as a maven project with all the dependencies added in pom.xml file. The Unit tests are
written using Junit framework which test few functions for their functionalities.
- Clone the repository or Download the zip file
- Opne projet using any IDE
- Build the project
- Look for the driver package
- Run the driver class
- Make appropriate choices
We have used Minst dataset which is in ubyte formate and to read the images in those formate we are using the
mnist-tools.jar.