Handwritten-Digit-Recognizer

This project uses the MNIST dataset to train a two layer neural network to reach an accuracy of 92.3% after 150 epochs. If you would like to run this project, download all the files and run the file titled: Project_005767973.m.

Name		Name	Last commit message	Last commit date
Project_005767973.m		Project_005767973.m
README.md		README.md
accuracy.m		accuracy.m
backward_propagation.m		backward_propagation.m
compute_cost.m		compute_cost.m
forward_propagation.m		forward_propagation.m
initialize_parameters.m		initialize_parameters.m
load_train_and_test_data.m		load_train_and_test_data.m
model_0.0010_2_150.png		model_0.0010_2_150.png
model_0.0100_2_100.png		model_0.0100_2_100.png
model_0.0100_2_150.png		model_0.0100_2_150.png
model_0.0100_2_50.png		model_0.0100_2_50.png
model_0.0100_3_150.png		model_0.0100_3_150.png
model_0.0100_5_150.png		model_0.0100_5_150.png
model_0.1000_2_150.png		model_0.1000_2_150.png
predict.m		predict.m
predict_single_image.m		predict_single_image.m
predict_single_image.png		predict_single_image.png
softmax.m		softmax.m
tanh2.m		tanh2.m
test_images.mat		test_images.mat
test_labels.mat		test_labels.mat
train_images.mat		train_images.mat
train_labels.mat		train_labels.mat
update_parameters.m		update_parameters.m
visualize_history.m		visualize_history.m

Navigation Menu

Repository files navigation

Handwritten-Digit-Recognizer

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

Handwritten-Digit-Recognizer

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages