Across explainable AI, game theory, digital forensics, biotechnology, and medical imaging.
An ongoing case study of read only, provenance preserving recovery of historical inode, namespace, extent, and payload evidence from a reformatted XFS volume.
→Co-authored work on detecting, tracking, and measuring king oyster mushroom mycelial aggregates in liquid culture. My role focused on AI and computer vision.
→A SHAP-guided approach to feature selection for ensemble classifiers, developed to improve classification while maintaining model transparency.
↗A study of early malware detection from partial Windows API-call sequences, comparing call-frequency and order-aware models across the first 5 to 100 calls.
Manuscript ↗Published in the ICCSS 2024 proceedings, pages 182–190, on identifying relevant and non-redundant features for classification.
Manuscript ↗ Full proceedings ↗A study of machine-learning methods for assisting breast-cancer detection from mammogram images.
↗