CNN Model from Scratch

This project is an implementation of a Convolutional Neural Network (CNN) from scratch. The purpose of this project is to learn and understand the details of building a CNN model without relying on existing libraries or frameworks. Table of Contents

• Introduction
• Features
• Installation
• Usage
• Examples
• Contributing
• License

Introduction

This project aims to provide a comprehensive understanding of Convolutional Neural Networks by building a CNN model from scratch. By implementing the various components of a CNN, including layers, neurons, and activation functions, we can delve into the inner workings of this powerful deep learning architecture.

Features

Implementation of a CNN model from scratch Customizable architecture with different layers, activation functions, and neuron configurations Forward and backward propagation algorithms for training the model Support for popular activation functions such as ReLU, sigmoid, and hyperbolic tangent

Installation

Clone the repository:

git clone https://github.com/Thiagovasc/cnn-model-from-scratch.git

Navigate to the project directory:

cd cnn-model-from-scratch

Install the required dependencies:

pip install -r requirements.txt

Usage

To use the CNN model from scratch, follow these steps:

Import the necessary classes and functions:

from models.cnn import CNN
from models.layer import Layer
from models.neuron import Neuron
from utils.train_utils import ActivationFunctions

License

This project is licensed under the MIT License.