Skip to content

Latest commit

 

History

History
88 lines (56 loc) · 4.08 KB

README.md

File metadata and controls

88 lines (56 loc) · 4.08 KB
Raw

Simple Neural Network from Scratch

Overview

This repository contains a Jupyter notebook where we build and explain a simple neural network from scratch using only basic Python. No external libraries like TensorFlow, PyTorch, or NumPy are used—just raw Python code to help beginners understand the fundamental concepts of neural networks.

What's Inside?

  • A complete explanation of forward propagation in neural networks.
  • A simple neural network model with:
    • 2 neurons in the input layer.
    • 2 neurons in the hidden layer.
    • 1 neuron in the output layer.
  • Detailed explanations of the role of weights, biases, and activation functions.
  • Step-by-step breakdown of how the network processes inputs to produce an output.

Project Goal

The goal of this project is to provide beginners with a clear and hands-on introduction to how neural networks work at a fundamental level. By the end of the notebook, you will understand:

  • What weights and biases are and how they influence a network’s output.
  • The importance of activation functions and how they introduce non-linearity.
  • How a network processes input data using forward propagation.

Setup Instructions

If you want to run the notebook on your local machine, follow the instructions below to set everything up.

Prerequisites

To run this notebook, you’ll need:

  • Python 3.x installed on your machine.
  • Jupyter Notebook installed.

If you don’t have Jupyter Notebook installed, follow these steps to set it up:

  1. Install Jupyter via pip:

    pip install notebook

  2. Clone the repository:

    git clone https://github.com/yourusername/your-repo-name.git
cd your-repo-name

  3. Launch Jupyter Notebook:

    jupyter notebook

  4. Open the notebook file (Forward-Propagation.ipynb) from the Jupyter interface and explore the code and explanations.

How the Neural Network Works

This neural network uses a basic architecture with one hidden layer to perform a simple task of binary classification. Here's a summary of the process:

  1. Input Layer: The input layer receives two inputs. These could be features like pixel values from an image or any other data.

  2. Hidden Layer: The hidden layer contains two neurons. Each neuron processes the input data by calculating a weighted sum of the inputs, adding a bias, and then applying an activation function (sigmoid) to determine its output.

  3. Output Layer: The output layer contains a single neuron that takes the outputs of the hidden layer, calculates a weighted sum, adds a bias, and applies the sigmoid activation function to produce the final prediction.

Forward Propagation

  • Step 1: Weighted Sum: Each neuron multiplies its inputs by its corresponding weights and calculates a weighted sum.
  • Step 2: Add Bias: A bias is added to the weighted sum, allowing the network to adjust its output.
  • Step 3: Apply Activation Function: The sigmoid activation function is applied to introduce non-linearity, which enables the network to learn complex patterns.
  • Step 4: Output: The output of the final neuron represents the network’s prediction, usually a value between 0 and 1, indicating the probability of a certain outcome.

Example Usage

If you run the notebook, you will see step-by-step explanations of how the network processes the input data. Here's a quick look at how you can run the notebook:

  • Modify the Inputs: Try changing the input values and observe how the network's output changes.
  • Explore Weights and Biases: Experiment with different weights and biases to see how they affect the output.
  • Understand Forward Propagation: Follow the detailed explanations in each cell to get a solid grasp of how forward propagation works.

Contributing

Contributions are welcome! If you have suggestions for improving this project, feel free to submit a pull request or open an issue. Together, we can make this repository even better for beginners.

License

This project is licensed under the MIT License - see the LICENSE file for details.