Skip to content

Latest commit

 

History

History
20 lines (15 loc) · 1.13 KB

README.md

File metadata and controls

20 lines (15 loc) · 1.13 KB

Handwritten Digit Recognition using FeedForward Neural Network

This Model Employs 'Feed-Forward Neural Network' architecture to identify digits in a handwritten digit
Data used here is sourced from MNIST Database

Final Accuracy: 96.65%

Dataset:

  • MNIST Handwritten Digits:
    Sample of Dataset

Accuracy & Loss:

  • Accuracy of the model plotted against no. of iterations (epochs)
    Accuracy vs. epochs
  • Loss of the model plotted against no. of iterations (epochs)
    Loss vs. epochs)

Predictions:

  • Correctly_predicted_ones Correctly_predicted_ones