Research

Current Research

University of Manitoba

My doctoral research is conducted at the Department of Economics, University of Manitoba.

My field paper focuses on Decision Theory, with particular emphasis on Random Expected Utility models.

Francophone Immigration Research

Université de Saint-Boniface

I am currently collaborating with Professor Faiçal Zellama, Professeur titulaire d'économie et de gestion at the Université de Saint-Boniface, on research examining francophone immigration in Manitoba.

Our work investigates issues of integration and systemic barriers faced by francophone immigrants in the province.

Working Paper: AI & Currency Detection

YOLO Banknote Detection

I am currently developing a working paper with Derrick Mirindi (University of Pennsylvania) on the application of YOLO Algorithm for banknote detection and authentication.

This research explores machine learning approaches to currency recognition and counterfeit detection systems.

Accepted — 2026 IEEE 3rd International Conference on Deep Learning and Computer Vision (IEEE DLCV 2026)

▶ Click here for Preliminary Results

20 USD — Detection Output

100 USD — Detection Output

The YOLOv8 model accurately identifies and localizes key security features on both the $20 and $100 banknotes, including serial numbers, Federal Reserve seals, and denomination markers, with high confidence scores (above 0.75).

YOLOv8s Eigen-CAM — Model Attention Heatmap

The Eigen-CAM activation maps reveal which regions of each banknote the model focuses on during inference. Warmer areas (red/yellow) indicate high attention zones, confirming that the model prioritizes security-relevant features such as watermarks, portraits, and denomination numerals rather than background textures.

Working Paper: Gaming Addiction & Sleep Quality

Multi-Dimensional Analysis

This paper presents a multi-dimensional analysis of the relationship between gaming addiction and sleep quality. We introduce a novel Game Addiction Risk Score (GARS), a composite index scaled from 0 to 100. We apply K-Means clustering to identify four distinct gamer profiles, fit polynomial regression models to capture non-linear trends, and test a mediation pathway through social isolation.

Our results show a strong negative correlation between daily gaming hours and sleep hours (r = −0.74, p < 0.001). The four identified clusters reveal distinct behavioral profiles with unique intervention implications.

Under review — IEEE Journal

▶ Click here for Preliminary Results

Figure 1 — Core Sleep & Gaming Analysis

The analysis reveals a strong negative correlation (r = −0.74) between gaming hours and sleep hours. GARS is significantly associated with lower sleep quality (ρ = −0.47, p < 0.001). One-way ANOVA confirms significant group differences across addiction risk levels (F = 254.0, p < 0.001).

Peer Reviewer Recognition

Recognized as a peer reviewer for Sustainable Environment, a journal published by Taylor & Francis / Routledge.

Reviewer Certificate — 2026