Data Scientist and ML researcher with a physics background. I build models, algorithms and data workflows that extract signal and insight from large, noisy and high-dimensional datasets.
I work at the intersection of data science, machine learning and computational science — turning raw measurements into models, decisions and scientific insight.
My work has involved millions of diffraction images, low-signal imaging, Bayesian analysis, unsupervised learning, deep neural networks, optimization and GPU/HPC workflows. The common thread: taking a difficult data problem and building a practical computational way through it.
Research-grade problems. Production-minded thinking.
Developed Bayesian and machine-learning approaches to classify, embed and characterize large-scale X-ray imaging datasets from EuXFEL experiments. Unsupervised learning revealed reaction trajectories and structural changes in biological samples.
Designed a derivative-free optimization algorithm for holographic single-particle imaging, robust to noisy nonlinear parameter spaces and extendable to heterogeneous nanoscale objects.
Built a cyclical deep-learning framework combining neural networks and genetic algorithms for inverse and forward design of nanophotonic metasurfaces.
Developed maximum-likelihood pattern-search methods for holographic single-particle imaging at XFELs, targeting weakly scattering nanoscale objects while improving robustness to noise, heterogeneity and computational scale.
Years of research have meant owning problems end-to-end: framing the question, wrangling data, designing algorithms, training models, running computation and communicating the result.
Computational Nanoscale Imaging · Data analysis · Machine learning · Scientific computing
Deep learning for inverse design of nanophotonic devices
Scattering theory and computational modeling
Statistical analysis, dimensionality reduction, feature extraction, visualization.
Deep learning, unsupervised learning, autoencoders, transfer learning.
Python pipelines, GPU acceleration, HPC, Slurm and scalable computation.
Optimization, probabilistic methods, noisy inverse problems and scientific modeling.
Peer-reviewed work spanning ML, imaging and computational design.
I'm interested in teams where rigorous data science, machine learning and computational thinking can create a real-world advantage.
abhishekmall101@gmail.com ↗