My Projects

PUBLICATION - Albatool Wazzan, Marcus Wright, Stephen MacNeil, and Richard Souvenir (2025). Evaluating the Impact of AI-Generated Visual Explanations on Decision-Making for Image Matching. 'In press', 30th International Conference on Intelligent User Interfaces (IUI '25), ACM. Conducted research for the NSF REU Site for Pervasive Computing at Temple University, within the Vision, Imaging, and Data Analysis Research (VIDAR) Lab under Dr. Richard Souvenir's mentorship. Focused on evaluating AI explanations, analyzing their impact on improving human interaction and decision-making with AI systems. Gained experience with Human-Computer Interaction, and Explainable AI (xAI) systems. Integrated AI explanation methods into a Deep Similarity Machine Learning Model and added to a web-based frontend to facilitate participant interaction with the AI explanations. Conducted a user study with modified frontend.

Designed and implemented advanced imputation strategies (mean/mode, predictive modeling, hybrid approach), handled outliers, and applied feature scaling to optimize model inputs. Reduced multicollinearity through correlation analysis, transformed cyclical variables with sine-cosine encoding, and mapped spatial data using latitude-longitude to enhance model interpretability. Evaluated multiple preprocessing pipelines across KNN, Logistic Regression, Decision Trees, and Random Forests, using statistical validation (ANOVA, Kruskal-Wallis) to assess their effect on predictive accuracy (~84-85%)

Built a comprehensive learning website designed to help first-generation, low-income students understand and navigate the world of credit cards. The platform offers a series of educational modules, each covering key aspects of credit. Integrated an AI Chatbot, designed to assist users by answering more detailed questions. Designed built-in content moderation for the chatbot by altering model context.

This personal website is a fully responsive and visually engaging platform built with Next.js and Tailwind CSS, designed to showcase projects, provide seamless navigation, and enhance user interaction. It features a dynamic project portfolio with smooth hover effects that reveal additional details, a sleek and fixed background for a polished aesthetic, and an intuitive contact form integrated with Slack webhooks for instant message delivery. The website is structured for scalability, with a modern navbar and footer, ensuring a professional yet personal touch. Designed with accessibility in mind, it offers a clean and modern color palette, a fluid layout, and interactive elements that enhance the browsing experience.