Neural-Network-from-Scratch

---

This directory is dedicated for the study of the book Neural Networks from Scratch (https://nnfs.io/) in which the author proposes to explain, both by coding and mathematically, the functioning of a Neural Network from within.

There is also a youtube playlist create by this same author, although incomplete till now.

It can be found at this link: https://www.youtube.com/watch?v=Wo5dMEP_BbI&list=PLQVvvaa0QuDcjD5BAw2DxE6OF2tius3V3

The book goes over the following topics:

1. Creating Neurons
2. Adding Layers
3. Creating Activation Functions
4. Calculating Error through Loss Functions
5. Optimizing - Derivatives
6. Optimizing - Gradient Descent, Partial Derivatives, Chain Rule
7. Backpropagation
8. Regulatization
9. Saving and Loading Models

In development:

---

Till now it has been documented all the code up to Optimizing classes of the book. There's left:

• Notebook with all optimizing classes
• Notebook with al regression classes
• Final notebook
• Final code (.py)
• .doc with mathematical explanation (Notion)
• .doc with Literature Review on Latests Advances in this topic (Notion)

Repo Structure

---

.
├── NN-Scratch-1            # Notebook with all classes up to Optimizing
├── NN-Scratch-2            # Notebook with all optimization classes
├── NN-Scratch-3            # Notebook with all Regression classes
├── NN-Scratch-4            # Final notebook
├── nn_scratch.py           # Final code
├── LICENSE
└── README.md