Development of a deep learning model to predict peak ground acceleration (and earthquake intensity) by analyzing the first 3 seconds of the p‑phase in seismograms under the supervision of Dr. Mohammed Eunus Ali. The model uses train‑time data augmentation to generate partial waves and employs a self‑supervised contrastive learning approach to align partial waves with the full waveform in the embedding space. To address the dataset's imbalance in large‑magnitude earthquakes, SMOTER (SMOTE for regression) and a density‑based sample loss weighing scheme were applied.

Dataset preparation of the physiological aspects and environmental factors of Bangladeshi Rickshaw Pullers. Supervised by Dr. A. B. M. Alim Al Islam. Collect skin temperature, electrodermal activity(EDA), pulse rate variability, EEG data, etc. as physiological aspects and environmental temperature, relative humidity, dust particle amount, pollution levels, etc. as environmental data. Accumulated data of over 50 rickshaw pullers, each having more than 20 minutes of data.

• Course Instructor: Introduction to Programming. Experienced in giving lectures about introductory Python.
• Course Instructor: Data Structure and Algorithms Sessional.
• Course Instructor: Structured Programming Language Sessional.

CGPA: 3.93/4.00
Notable Courses: Machine Learning, Simulation and Modelling, Computer Graphics, Artificial Intelligence, Data Structures and Algorithms, Computer Architecture, Database Management Systems, Software Engineering, Operating Systems, Computer Networks, Theory of Computation, Numerical Methods, Discrete Mathematics.