Skip to content

README.md

Latest commit

 

History

History
19 lines (14 loc) · 681 Bytes

README.md

File metadata and controls

19 lines (14 loc) · 681 Bytes

MNIST Hand Written Digit Classifer

Back propagation feed neural network that classifies hand written digits in C++ with a sucess of 90%>

Class Project for CSCI:360 at USC

Description
Task1: Simple perceptron that learns XOR table
Task2: Used for analysis with different hyperparamters
Task3: Back propagation feed neural network

How to Run Task3

  • cd task3
  • ./task3

Output for task3

  • First part is training the date with sigmoid activation function
    It will take 15 epochs to train the model then it will produce the success validation results after training

  • Second part is using a different activation function which is tanh.