Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 429 Bytes

File metadata and controls

13 lines (10 loc) · 429 Bytes

handwritten-character-recognition

A piece of software to recognize handwritten capital characters.

USAGE

Explanations are in turkish ( as this is a project I did in university )

Link of dataset is in handwritten-az folder. After downloading the dataset

  1. run import_dataset.m
  2. run create_network.m
  3. after training run test_network
  4. view_matrix.m is intended to see network's parameters.