My research applies quality control and validation methodologies to predictive models, especially machine learning systems. My current work is focused towards Explainable AI (XAI) and adversarial machine learning, investigating how models behave, degrade, and how they can be made more resilient to both naturally occurring and deliberately induced error. Some of the recent domains I have done work in are detailed below. 

Adversarial Resilience & Cybersecurity

• Malware Detection & Attribution: Utilizing signal processing methods to identify, classify, and trace malicious payloads.

• Physical-Domain Security: Evaluating adversarial vulnerabilities using non-visual spectrum data, with a focus on Digital Night Vision (Infrared) operating environments.

Explainable AI (XAI) & Algorithmic Transparency

• Natural Language Processing (NLP): Extracting reliable, quantitative metrics from highly unstructured text data.

• Large Language Models (LLMs): Investigating structural transparency and emergent behaviors, including mechanisms for reliable self-output detection and validation.

• Calibration and Validation of XAI Methods: Identifying and evaluating limitations in the practical effectiveness of XAI methods.

Statistical Quality Control & Systems Evaluation

• Algorithmic Quality Control: Integrating modern probabilistic methods, such as Conformal Prediction, into standard systems engineering and monitoring frameworks.

• Educational Systems Evaluation: Applying predictive modeling to measure instructional effectiveness.

Data Analysis & Applied Statistics

• Experimental Design & Inference: Evaluating research methodologies, constructing validation frameworks, and performing data analysis to ensure the integrity of third-party research.

• Applied Probability & Risk Modeling: Developing quantitative risk assessments and probabilistic models.

Courses