Neural Nets and Deep Learning

This library is used for learning about neural networks. Reading material is from Neural Nets and Deep Learning and original source code is from this repository.

Pt. 1: Reading Textbook

Pt. 2: Implement Neural Network

Pt. 3a: Confusion Matrix

• Code for the confusion matrix can be found here in the evalution function.

Pt. 3b: Visualize Incorrect Predictions

• Code to visualize incorrect predictions can be found here in the transform_digit function.
• For example, this number is a 9 but was predicted as a 7:

Pt. 3c: Save and Load Models

• Code to save and load pretrained models can be found here in the save_network and load_network functions.
• The /src/pretrained_models folder contains some pretrained models used during this project.

Pt. 4: ROC Curves

• The code to implement an ROC curve can be found here in the roc function.

Pt. 5: Hyperparameter Experimentation

• The experiments and their results can be found here.