6 papers
Cosmo3DFlow: Wavelet Flow Matching for Spatial-to-Spectral Compression in Reconstructing the Early Universe
Md. Khairul Islam, Zeyu Xia, Ryan Goudjil +3
Reconstructing the early universe from the evolved present-day universe is a challenging and computationally demanding problem in modern astrophysics. We devise a novel generative…
OmniSpectra: A Unified Foundation Model for Native Resolution Astronomical Spectra
Md Khairul Islam, Judy Fox
We present OmniSpectra, the first native-resolution foundation model for astronomy spectra. Unlike traditional models, which are limited to fixed-length input sizes or configuratio…
Scalable Cosmic AI Inference using Cloud Serverless Computing
Mills Staylor, Amirreza Dolatpour Fathkouhi, Md Khairul Islam +4
Large-scale astronomical image data processing and prediction are essential for astronomers, providing crucial insights into celestial objects, the universe's history, and its evol…
WinTSR: A Windowed Temporal Saliency Rescaling Method for Interpreting Time Series Deep Learning Models
Md. Khairul Islam, Judy Fox
Interpreting complex time series forecasting models is challenging due to the temporal dependencies between time steps and the dynamic relevance of input features over time. Existi…
Large Language Models for Financial Aid in Financial Time-series Forecasting
Md Khairul Islam, Ayush Karmacharya, Timothy Sue +1
Considering the difficulty of financial time series forecasting in financial aid, much of the current research focuses on leveraging big data analytics in financial services. One m…
Does Differential Privacy Impact Bias in Pretrained NLP Models?
Md. Khairul Islam, Andrew Wang, Tianhao Wang +3
Differential privacy (DP) is applied when fine-tuning pre-trained large language models (LLMs) to limit leakage of training examples. While most DP research has focused on improvin…