AstroPhotoZ: Photometric Redshift Estimation
A comprehensive framework for estimating photometric redshifts using Gaussian processes and machine learning techniques, validated on Stripe 82X before applying to wide X-ray fields.
GitHubMy research focuses on the intersection of observation, computation, and theory. My current and ongoing research projects are:
Testing Modified Gravity theories via galaxy clustering and CMB lensing data
This project forecasts signatures of modified gravity (including f(R) and DGP gravity) using cross-correlation analyses between galaxy clustering and CMB lensing. Incorporates simulation and survey specifications (LSST/DESC, Simons Observatory), Fisher matrix forecasting, MCMC parameter inference, and fast emulators for theoretical predictions.
Supervisors: Dr. Tanveer Karim; Dr. Georgios Valogiannis
Machine Learning for High-Energy Physics Jet Classification
The Jet-Tagging-Experiments repository develops an explainable Particle Chebyshev Network that uses physics-motivated interaction features (ln Δ, ln kT, ln z, ln m²) to classify jets on CERN Open Data (JetClass), advancing QCD-informed machine learning for particle physics.
Supervisors: Dr. AKM Mahbubur Rahman; Dr. Amin Ahsan Ali; Dr. Arshad Momen (PI)
Physics-Informed Neural Networks for Cosmological Inflation (Exploratory Work)
Explores the application of Physics-Informed Neural Networks (PINNs) to model stochastic inflation in early universe cosmology. This exploratory work investigates how machine learning techniques can be integrated with physical constraints to solve differential equations governing inflationary dynamics.
Supervisor: Dr. Rafid Mahbub
Machine Learning for GRB Data & Redshift Estimation
Contains tools and experiments focused on using machine learning to estimate redshifts and classify gamma-ray bursts using Fermi and other astrophysical data. Explores advanced algorithms and astrophysical feature engineering for transient phenomena.
Supervisors: Dr. Soebur Razzaque; Dr. Saeeda Sajjad
Radha Gobinda Chandra, a Radio Galaxy Classifier
An advanced machine learning-based classifier designed to identify and categorize radio galaxies from astronomical survey data. This tool leverages deep learning techniques to automatically classify different types of radio sources, significantly improving the efficiency of radio astronomy research.
Supervisors: Dr. Khan Muhammad Bin Asad, Dr. AKM Mahbubur Rahman; Dr. Amin Ahsan Ali, Dr. Arshad Momen (PI)
Radio Frequency Interference Identification and Elimination
This research is on identifying and eliminating radio frequency interference in astronomical observations. It addresses one of the most critical challenges in radio astronomy, where interference from human-made sources can significantly impact the quality of astronomical data.
Supervisors: Dr. Khan Muhammad Bin Asad, Dr. AKM Mahbubur Rahman, Dr. Arshad Momen (PI)
Check my projects in astrophysics, machine learning, and interdisciplinary data analysis:
A comprehensive framework for estimating photometric redshifts using Gaussian processes and machine learning techniques, validated on Stripe 82X before applying to wide X-ray fields.
GitHubRecognizing normal vs. unusual activities related to Parkinson's using multimodal sensor data from wearables, classifying ten distinct activity classes for early diagnosis.
GitHubCurating and managing radio galaxy datasets from the FIRST survey with labeled morphological classifications (FRI, FRII, compact, bent) for ML training.
GitHubAn ML pipeline for detecting exoplanets from Kepler and TESS data using logistic regression, KNN, and random forest with data augmentation techniques.
GitHubLeveraging ML to predict semiconductor band gaps from crystal structure data using Pymatgen and Matminer for optoelectronics and photovoltaics discovery.
GitHubA signal processing and ML framework to detect and mitigate RFI in radio astronomy, integrating traditional techniques with deep learning including GANs.
GitHubA comprehensive toolkit for managing petabyte-scale astronomical datasets through data mining, ML, and statistical methods with Jupyter notebooks and real SDSS data.
GitHubExploring Neural ODEs for physics and astrophysics — continuous-depth neural networks modeling stellar evolution, gravitational dynamics, and time-dependent phenomena.
GitHubEducational tutorials for radio astronomy data analysis using radio-astro-tools and Astropy — spectral cubes, PV diagrams, moment maps, and parallel processing with Dask.
GitHubA curated guide to essential astronomical tools, datasets, and learning materials across various subdisciplines of astronomy for students and researchers.
GitHubLecture notes, problem sets, and code implementations covering the CMB, large-scale structure formation, and other fundamental topics in modern cosmology.
GitHubComparative study of denoising techniques — classical signal processing vs modern ML approaches — with benchmarks for time-series and spectroscopic observations.
GitHubMonte Carlo simulations of radioactive decay chains in Cython, benchmarked against Numba, Pybind11, C++, Julia, and MATLAB for performance analysis.
GitHub