Neural Network Using Numpy

Introduction

Welcome to the Neural Network Using Numpy project! 🎉 In this assignment, you'll dive into the fascinating world of neural networks by building one from scratch using just Numpy. Your goal is to create a network that can classify handwritten digits (0-9) from the MNIST dataset.

Table of Contents

• Introduction
• Project Overview
• Sections Covered
• Requirements
• Getting Started
• Model Training and Predictions
• Resources
• Jupyter Notebook Viewer

Project Overview

In this project, you will:

• Implement the essential steps to build a neural network.
• Perform feedforward operations.
• Compute loss functions.
• Execute backpropagation for optimization.
• Update parameters (weights and biases).
• Train your model to classify handwritten digits.

Sections Covered

1. Data Preparation: Load and preprocess the MNIST dataset.
2. Feedforward: Implement the forward pass of the network.
3. Loss Computation: Calculate the loss to measure the accuracy of the network.
4. Backpropagation: Compute gradients to optimize the network.
5. Parameter Updates: Adjust the weights and biases to improve model performance.
6. Model Training and Predictions: Train the neural network and make predictions on new data.

Requirements

Before you begin, ensure you have the following installed:

• Python 3.x
• Numpy
• Matplotlib (optional, for visualization)

Getting Started

To get started with the project:

1. Clone the repository:

   git clone https://github.com/AbhinavSharma07/Neural-Network-Using_NUMPY.git

2. Navigate to the project directory:

   cd Neural-Network-using-Numpy

3. Install the required dependencies:

   pip install -r requirements.txt

4. Load the dataset:

   • You can refer to the mnist.pkl.gz file for the dataset.
   • The file will be automatically loaded in the Jupyter notebook.

5. Open the Jupyter Notebook to start coding:

   jupyter notebook Neural-Network-using-Numpy.ipynb

Model Training and Predictions

The notebook is designed to guide you through the entire process of building and training a neural network. By the end of this project, you'll have a fully functional model capable of recognizing handwritten digits with high accuracy.

Resources

• MNIST Dataset: The classic dataset for handwritten digit recognition.
• Numpy Documentation: Numpy Official Documentation.
• Neural Networks: Understanding how neural networks work and their applications.