Enhanced-Classification-of-Class-Imbalanced-Datasets-using-Hybrid-Quantum-Models

This project explores hybrid quantum-classical models for enhanced classification of class-imbalanced datasets. It implements Hybrid Quantum Neural Networks (HQNNs) and quantum-enhanced algorithms (Random Forest, KNN, SVM) to improve minority class prediction.

Project Overview

Class imbalance is a common problem in real-world datasets where one class is significantly less represented than others. This project investigates the use of hybrid quantum-classical models to address this challenge, aiming to improve the prediction of minority classes.

Key Components

The project is divided into two main parts:

1. Hybrid Quantum Neural Networks (HQNNs):

   • Explores various HQNN architectures for classification.
   • Utilizes Qiskit to implement quantum circuits within neural networks.
   • Applies HQNNs to datasets like Credit Card Fraud, Melanoma Lesion Classification, Higgs Boson, and Breast Cancer Wisconsin.
   • Focuses on techniques to integrate classical neural networks with quantum circuits for enhanced feature extraction and classification.

2. Quantum-Enhanced Classical Algorithms:

   • Implements quantum versions of Random Forest, KNN, and SVM.
   • Uses quantum circuits to compute kernel matrices or embed data.
   • Evaluates these models on the Credit Card Fraud and Breast Cancer datasets.
   • Compares the performance of these quantum-enhanced models with their classical counterparts.

Datasets

The following datasets were used in this project:

• Credit Card Fraud Detection dataset
• Melanoma Lesion Classification dataset
• Higgs Boson dataset
• Breast Cancer Wisconsin (Diagnostic) dataset

Key Technologies

• Qiskit
• PennyLane
• scikit-learn
• PyTorch

Authors

• B.S. Tejas (ED22B004) - HQNN part
• Kanishq Garg (ED22B051) - Quantum-Enhanced Classical Algorithms part

Course

This project was completed for the course DA6300 - Quantum Computing And Machine Learning at IIT Madras.

Repository Contents

• HQNN.ipynb - Jupyter Notebook for the HQNN part of the project.
• qcml.ipynb - Jupyter Notebook for the Quantum-Enhanced Classical Algorithms part of the project.

Usage

Instructions on setting up the environment and running the code can be found within the individual notebook files.