Improving Reinforcement Learning Sample-Efficiency using Local Approximation
Preprint
CV
Downloadable CV available here.
Education
- Ph.D in Computer Science, Nanyang Technological University, 2025 (expected)
- Thesis in probabilistic guarantees for reinforcement learning
- B.Eng. in Computer Science, Honours (Distinction), Nanyang Technological University, 2021
- Specializations in Data Science and Artificial Intelligence
Work experience
- Aug 2019 –: ML Researcher
- Nanyang Technological University
- Duties includes: scientific research and teaching assistant work
- Notable achievements: see publications, talks and teaching
- May 2021 – Oct 2021: Data Scientist (intern)
- Munich Reinsurance
- Duties included: data cleaning, consulting, building ML pipelines for fraud detection
- Notable achievements: built pricing and risk models using geospatial data
- Aug 2020 – Dec 2020: Data Scientist/Software Engineer (intern)
- ST Engineering Land Systems
- Duties included: experimenting on image processing and building simulation tools
- Notable achievements: developed open source threading and parsing libraries
Skills
Programming (C++, Python, Java),Linux,Quantitative Analysis,Research,Machine Learning
Publications
Guaranteeing Out-Of-Distribution Detection in Deep RL via Transition Estimation
Prashant, Mohit, et al. "Guaranteeing Out-Of-Distribution Detection in Deep RL via Transition Estimation". Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 39. No. 12. 2025.
PAC-Based Formal Verification for Out-of-Distribution Data Detection
Prashant, Mohit, and Arvind Easwaran. "PAC-Based Formal Verification for Out-of-Distribution Data Detection." 2022 6th International Conference on System Reliability and Safety (ICSRS). IEEE, 2022.
Generative Adversarial Masking (Student Paper)
Prashant, Mohit. "Generative Adversarial Masking" International Student Conference on Artificial Intelligence. 2021.
Talks
Addressing Guaranteed Out-Of-Distribution Policy Execution
Talk at 3 Minute Thesis, Singapore
Efficiently Scaling Model-Free Reinforcement Learning with Local Approximation
Talk at University of Pennsylvania, SEAS, Philadelphia, PA, USA
Building Robust Systems Using Adversarial Input Masking
Talk at International Student Conference on Artificial Intelligence, Singapore
Teaching
Service and leadership
- Served as reviewer in
AAAI,ICML,IJCAI,AAMAS,Transactions in AIand others - Red Cross Singapore Volunteer, 2020-
- Red Cross Youth Officer, 2018-2019