Research Publications

Thesis

Quantum Kernel Techniques for Optimizing Support Vector Machine Performance
This thesis work explores the integration of quantum computing with classical SVMs through quantum-enhanced kernel functions. By utilizing quantum feature maps, quantum phase estimation accelerates kernel computations and improves high-dimensional data separability. It addresses scalability challenges, enhances convergence rates, and optimizes performance in complex, non-linear classification tasks, pushing the boundaries of machine learning with quantum innovations.

Ongoing Research

  • Automated Recognition of Fundamental Human Movements and Postures
  • Development of Self-Navigating Systems for Chittagong’s Roadways
  • Development and Implementation of Meteorological Monitoring Network for Chittagong
  • Time Series Analysis for Enhanced Long-Term Accuracy in Localized Weather Forecasting
  • MRI-Based Image Segmentation
  • Removing Hair and Filling Image Gaps with Vision Transformer Models
  • Quantum Machine Learning Architecture for Handling Skewed Datasets

Personal Projects

  • MNIST image classification with QNN
    Tools: PyTorch, Classiq SDK
    Used quantum technology from Classiq to recognize digits through hybrid QNN.
  • Health Bee
    Tools: Android Studio, CI/CD utility
    Basic performing mobile app developed following SE norms and methods, with proper documentation.
  • Website for Bit2Byte
    Tools: JavaScript, C#, HTML, CSS
    Website with admin panel, user panel, newsletter box, and basic website features.
  • Introduction to Quantum Chemistry Notebook
    Tools: Qiskit, Cirq
    Tutorial notebook on mapping, encoding, VQE, and QPE in quantum chemistry.
  • Piksel
    Tools: Android Studio, Java SDK
    Android app where artists can auction artworks and clients can purchase through bidding.
  • Solution to Pennylane Codebook
    Tools: Pennylane
    Solution repository for tasks in the PennyLane codebook, from different subdomains of quantum computing.