This repository contains the source code for various types of Neural Networks from scratch, as well as their applications. The types of Neural Networks in this repository range from simple neural networks (single layer, hidden layer) to more advanced multi-layer neural networks. The source code in this directory is Python and no external libraries libraries like Scikit-learn, Keras etc. are used.
This Directory contains the source code for some of the basic types of Neural Networks, these include:
- Single Layer (Multi-input -> Single-output): This directory contains source code for a general single layer n-to-1 neural network.
- Single Layer (Multi-input -> Multi-output): This directory contains source code for a general single layer n-to-m neural network.
- Hidden Layer (Multi-input -> Single-output): This directory contains source code for a general non-linearity (hidden layer) n-to-1 neural network.
- Hidden Layer (Multi-input -> Multi-output): This directory contains source code for a general non-linearity (hidden layer) n-to-m neural network.
This directory uses the hidden layer code from the Simple Neural Networks directory to make different types of Perceptrons, which include:
- Single Layer Perceptron: This directory contains the source code for a single binary perceptron neural network.
- Multi-Layer Perceptron: This directory contains the source code for a multi-layer activation perceptron neural network.
This directory contains source code for a simple neural network that can be used to differentiate between malignant and benign tumors in patients. This is achieved by using a Multi-Layer Neural Network with Sigmoid and Threshold non-linearities. This directory contains:
- Multi-Layer Neural Network: This directory contains the multi-layer neural network code that I will use.
- Diagnosis Code: This directory contains the code for the diagnosis using the afforementioned Neural Network.
Please feel free to reach out to me using:
rohan.b.singh54@gmail.com
rxs1182@case.edu