Skip to content

ReadMe.md

Latest commit

 

History

History
106 lines (82 loc) · 3.95 KB

ReadMe.md

File metadata and controls

106 lines (82 loc) · 3.95 KB

Blood Cell Cancer Classification using CNN and EfficientNetB3

License Status Contributions

Table of Contents

Overview

This project aims to classify blood cell images to detect cancerous cells using deep learning techniques, specifically Convolutional Neural Networks (CNN) and EfficientNetB3 architecture. The goal is to develop a robust model that can accurately differentiate between cancerous and non-cancerous blood cells.

Dataset

The dataset used in this project is sourced from Kaggle, containing images of various blood cell types. The dataset is organized into folders for each cell type.

Installation

Ensure you have the following libraries installed:

  • TensorFlow
  • Keras
  • NumPy
  • Pandas
  • Matplotlib
  • Seaborn
  • OpenCV
  • PIL

Install the required libraries using:

pip install tensorflow keras numpy pandas matplotlib seaborn opencv-python pillow

Getting Started

  1. Clone the repository: 
    git clone https://github.com/recodehive/machine-learning-repos.git
  2. Navigate to the project directory: 
    cd machine-learning-repos/Detection Models/Blood Cell Cancer Detection using CNN and EfficientNetB3
  3. Ensure you have the required libraries installed as mentioned in the Installation section.

Notebook Structure

Import Necessary Libraries

This section imports all the required libraries for data handling, preprocessing, and building the CNN model.

Reading the Data

The data is read from the specified directory, and file paths along with labels are stored in a DataFrame.

Explore the Data

Exploratory data analysis is performed to understand the distribution and characteristics of the dataset.

Data Preprocessing

Data preprocessing steps include splitting the dataset into training and validation sets and augmenting the images.

Building the Model

A Convolutional Neural Network (CNN) model is built using Keras. The model architecture includes convolutional layers, pooling layers, and dense layers.

Training the Model

The model is trained on the preprocessed dataset with specified parameters.

Evaluating the Model

Model performance is evaluated using metrics such as confusion matrix and classification report.

Using EfficientNetB3

EfficientNetB3 architecture is used to enhance the model's accuracy. The pre-trained EfficientNetB3 model is fine-tuned on the dataset.

Conclusion

Summary of the findings and results, including insights on model performance and potential improvements.

Results

The project demonstrates the capability of CNN and EfficientNetB3 in classifying blood cell images with high accuracy. The final model achieved an accuracy of XX% on the validation set.

Future Work

  • Explore the use of other pre-trained models.
  • Implement more advanced data augmentation techniques.
  • Deploy the model as a web application for real-time predictions.

Authors

License

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

You can copy and paste this improved version into your ReadMe.md file.