Academic Publications

ADAPT: Asymmetric Device Adaptive Parallel Training for Small-Scale Heterogeneous Systems

Thesis Research
Masters Thesis / MPS Lab • 2024 - 2025

Investigated asymmetric parallel training strategies spanning model, pipeline, and data distribution formats on heterogeneous GPU environments. Formulated dynamic scheduling costs to minimize cross-device latency.

Detection of Malicious URLs Through an Ensemble of Machine Learning Techniques

Cited by 15
Published in IEEE-CSDE Conference Proceedings • 2021

Formulated a cascading, high-accuracy url classification schema. Deployed 6 independent attribute-specific feature classifiers passing logical inputs to a final ensemble decision tree.

Movie Success Prediction and Performance Comparison Using Various Statistical Approaches

Cited by 3
Published in MLDS Conference Proceedings • 2021

Evaluated and compared mathematical accuracies of distinct classifiers (regression models, neural layers, decision trees) predicting financial yield success metrics based on production features.

A Comprehensive Study on Various Statistical Techniques for Prediction of Movie Success

Cited by 3
Published in International Journal of Artificial Intelligence & Applications (IJAIA) • 2021

Conducted extensive correlation audits and feature importance mappings to analyze box office predictability metrics based on cast strength, release windows, and budget ratios.

Research Interests & Directions

Machine Learning Accelerators

Exploring specialized hardware compilers, scheduling layers, and performance optimizations for large language model workloads.

Symmetric & Asymmetric SMPC

Developing Secure Multi-Party Computation protocols enabling collaborative machine learning over encrypted, privacy-safe pipelines.

Quantum Computing Simulators

Researching NISQ quantum hardware limitations, state vector simulations, and algorithms under the guidance of MPS lab collaborators.