pp_conv.png

I am a PhD student in the Department of Computer Science at Purdue University, fortunate to be advised by Drs. Ming Yin and Rajiv Khanna.

I am broadly interested in Human-Centered AI/ML, drawing insights from not only machine learning but also human-computer interaction, behavioral sciences, crowdsourcing etc. My current research focus is on Human-AI Interaction and eXplainable AI (XAI). Specifically, I am interested in building AI systems that complement human capabilities by design and exploring ways to enhance the practical utility of explainability techniques, ultimately aiming to elevate the performance of human-AI teams.

Before joining Purdue, I earned my undergraduate degree from LUMS and had the privilege of collaborating with researchers from TUKL-NUST R&D Lab, University of Virginia and University of Notre Dame. I also briefly wore a Data Scientist hat at Afiniti, where I contributed to refining Customer Experience Artificial Intelligence (CXAI) by optimizing agent-caller pairing in call center-like environments.

News

Dec 2023 Attending NeurIPS to present work on Support Vector Effect in DNNs.
Oct 2020 Work done as undergraduate thesis now accepted as a full paper at IEEE ISI 2020. Excited to attend my first conference.
Sep 2020 Paper on using deep generative models to enhance predictive performance accepted at IEEE ICDM Workshops. Looking forward to present it in Italy!

Publications

On the Support Vector Effect in DNNs: Rethinking Last Layer Sensitivity-based Instance Attribution

Hasan Amin and Rajiv Khanna
Attributing Model Behavior at Scale Workshop at NeurIPS, New Orleans, December 2023.

Give Weight to Human Reactions: Optimizing Complementary AI in Practical Human-AI Teams

Hasan Amin, Zhuoran Lu, and Ming Yin
Artificial Intelligence & Human Computer Interaction Workshop at ICML, Hawaii, July 2023.

Using Deep Generative Models to Boost Forecasting

Hasan Amin and Ahmed Abbasi
IEEE International Conference on Data Mining (ICDM) Workshops, Sorrento, Italy, November 2020.

Phishcasting: Deep Learning for Time Series Forecasting of Phishing Attacks

Hasan Amin, Mustafa Abbasi, and Ahmed Abbasi.
IEEE International Conference on Intelligence & Security Informatics, Virtual, November 2020.