Md Mehedi Hasan

Bachelor of Science in Computer Science and Engineering (CSE)

Bangladesh University, Dhaka

About Me-

I aim to pursue a Ph.D. in Computer Science, driven by my passion for tackling complex systemic challenges through data structures and algorithmic problem-solving techniques, a journey that began early in my undergraduate studies.

My academic achievements include a CGPA of 3.91/4.00, ranking 1st in my class at Bangladesh University, Dhaka. My research has advanced CNN and attention mechanisms like CNN+CBAM for improved data classification accuracy and developed predictive models for customer satisfaction. Additionally, I contributed to optimizing PCF-SPR biosensors by predicting confinement loss using gradient boosting regression.

Education

➢Bachelor of Science in Computer Science and Engineering (2020-2023) | 1st Position

➢Higher School Certificate-2015

➢Secondary School Certificate 2013

Training

➢Introduction to Java Programming: Java Fundamental Concepts — Coursera

➢C for Everyone: Programming Fundamentals

➢Web Development (MERN) — Programming Hero

Research Interest

➢Machine Learning for Predictive Analytics

➢Deep Learning

➢AI for Human-Computer Interaction

My Expertise

I have been passionate about problem-solving from a young age and specialize in leveraging CNN, CBAM, and deep learning techniques to address complex challenges. My previous work has focused on image classification using machine learning. Beyond academics, I actively engage in exploring problem-solving strategies for data structures and algorithms. I continually seek innovative approaches to evaluate machine learning models and uncover how they adapt to various patterns and structures.

Skills

Programming Languages: Python, Java, C, C++
Web Development: ReactJS, Node.js, MongoDB, Express, Firebase
Machine Learning Tools: TensorFlow, Keras, Scikit-learn
Algorithms: SVM, CNN, ANN, GAN, RNN, KNN
Data Analytics: Pandas, Numpy, Matplotlib

My Experience

Industry Experience

Inovetix (Sept 2023 - Nov 2024)

Designed machine learning models to predict customer satisfaction, driving strategic business decisions.
Developed interactive dashboards for actionable insights from large datasets.

Research Assistant

Bangadesh University, Dhaka (Dec 2022 - Dec 2024)

Developed predictive frameworks integrating machine learning algorithms for healthcare applications.
Explored the optimization of PCF-SPR biosensors using AI techniques, achieving significant advancements in predictive accuracy.

My Research

Prediction of Confinement Loss for PCF-SPR Biosensor Utilizing Machine Learning: This research represents the development of a novel Photonic Crystal Fiber (PCF) biosensor with Surface Plasmon Resonance (SPR) to predict confinement loss, a critical factor impacting sensitivity in biosensors used for biotechnology and medical diagnostics. Created a dataset with key features, including wavelength, effective real and imaginary parts, and confinement loss, to train machine learning models. Gradient Boosting Regression (GBR) was trained and compared with other models, demonstrating superior performance in metrics like Mean Squared Error (MSE) and Signal-to-Noise Ratio (SNR). This work provided valuable insights for enhancing the sensitivity of bio-sensing devices.

Improving Esophageal Cancer Diagnosis: A Novel CNN+CBAM Architecture with GAN-Enhanced Data and Smooth Grad-CAM++ Interpretability: Proposed a CNN+CBAM architecture with GAN-enhanced data and Smooth Grad-CAM++ interpretability. Achieved 99.8% accuracy, 99.84% precision, 99.7% recall, and 99.8% F1-score, outperforming other models such as CNN and LSTM. This work offers a robust and interpretable solution for early-stage cancer detection with potential for clinical use.

MEHEDI