Research

Undergraduate Thesis

Thesis Title: “A Decentralized Learning-Based Approach to Classify Colorectal Cancer Using Deep Learning Leveraging XAI”

Goal: To create a decentralized image classification deep learning model for medical image classification using large-scale cancer histopathological tissue slide images. Our research was successful to obtain higher performance metrics on federated learning models. Ans was later submitted for conference paper.

Supervisor: Dr. Md. Ashraful Alam, Assistant Professor, Brac University.

Co-Supervisor: Md. Tanzim Reza, Lecturer, Brac University.

Publication

H. H. Sabuj, P. R. Gomez, K. E. Mubin, A. Hossain, S. Y. Seaum, M. T. Reza, M. G. R. Alam “A Comparative Study of Machine Learning Classifiers for Speaker’s Accent Recognition” IEEE World AI IoT Congress 2023(IEEE-AIIoT 2023)

DOI: 10.1109/AIIoT58121.2023.10174511

N. Arthi, K. E. Mubin, J. Rahman, M. Rafi, T. T. Sheja, M. T. Reza, M. A. Alam “Decentralized Federated Learning and Deep Learning Leveraging XAI-Based Approach to Classify Colorectal Cancer” The 9th IEEE Asia-Pacific Conference on Computer Science and Data Engineering 2022 (IEEE-CSDE 2022)

DOI: 10.1109/CSDE56538.2022.10089344

Ongoing Research

Title: "An Approach for Forensic Numerical Handwriting Analysis Using Multi-Tasking Convolutional Neural Network"

Goal: To develop a multi-tasking model capable of categorizing male/female handwriting as well as numerics. Research is in the early stage, initial implementation is completed but failed to achieve desired results.

Page updated

Google Sites

Report abuse