My work in AI-CARING focused on investigating the adjustment of trust among older adults towards robots that exhibit affiliation changes in a human-robot collaboration context.

• Human-Robot Interaction
• Human-Computer Interaction
• Python
• PyQt

I worked under the supervision of Professor Reza Ahmadzadeh to investigate the multi-agent multi-armed bandit problem in reinforcement learning. The main goal was to expedite the emergence of consensus among the agents in a network faced with a decision making problem. I worked on the implementation of multi-agent multi-armed bandit algorithms and simulations to test the performance of various optimization methods. The culmination of this work is my master's thesis, which focuses on accounting for various real-world constraints in the multi-agent multi-armed bandit context such as agent competency, bandit difficulty, and sequences of bandits, and optimizing team performance in the presence of these factors.

• Python
• NumPy
• Numba
• Matplotlib
• NetworkX

My research in the ARL STRONG project focused on developing a web interface to conduct virtual human-robot trust experiments. I created a web application that simulates a 2D human-robot collaboration game that examines the development of trust in human-robot networks by collecting various responses during the participants' playtime. So far, I deployed over 30 variations of this game on Amazon MTurk and Prolific which allowed hundreds of participants to contribute to the study's dataset. Several papers have been published based on the results and some are currently under review.

• HTML/CSS/JS
• jQuery
• Python
• Express
• MongoDB
• Heroku