Neural Networks Tutorials

In this repo, we have implemented 3 different neural nets in order to showcase how they work. each implementaiton is well documented to serve as a stand alone tutorial. Each implementation designed to showcase different case of implementation, where one implemented from scratch without using any library, another one implemented with tensorflow, and the third one used a pre trained model to solve a similar problem.We will used Python programing language in those implementations, with Jupyter IDE Notebooks as source code.

Enviroment Preperations

• Python 3: Follow the guide in this link to install Python 3 on Ubuntu or any other linux distribution and this link for Mack.

• Anaconda: Download and install from here.

• Jupyter IDE: it should be installed automatically with Anconda, in order to run the notebook, write this command in ther terminal.

  jupyter notebook
  

  Congratulations, you are now ready to go 👏.

Folder Structures

• Neural Network from scratch (Hotdog Classification): Network built from scratch.
• californiaHousePricesPrediction: Using tensorflow.
• Dog Breed Image Classification (Using ResNet50).ipynb: Using pre-trained model.

Neural Nets FAQ

We have included a very popular and useful neural net FAQ done by faqs.orh that well document important questions about neural nets, you can find the FAQs in this folder