My research is on problem-independent algorithms for sequential decision problems. I have a specific interest in algorithm selection for resource-constrained environments. I investigate how to compare the performance of nature-inspired learning techniques (mostly reinforcement learning & evolutionary computing for now) for dynamic optimisation.
Member of ALICE at Aston University.
Aman Soni, Peter R. Lewis, and Anikó Ekárt. "Offline and Online Time in Sequential Decision-Making Problems", Proceedings of the IEEE Symposium Series on Computational Intelligence 2016. View Cite this paper:
@inproceedings{soni2016offline,title={Offline and online time in Sequential Decision-Making Problems},author={Soni, Aman and Lewis, Peter R and Ek{\'a}rt, Anik{\'o}},booktitle={Computational Intelligence (SSCI), 2016 IEEE Symposium Series on},pages={1--8},year={2016},organization={IEEE}}
Aman Soni, Peter R. Lewis, and Anikó Ekárt. “Synergies Between Reinforcement Learning and Evolutionary Dynamic Optimisation.” In: Artificial Life and Intelligent Agents. Springer International Publishing, 2016. [to appear]
Talks:
Gaia chatbot for mental health self-assessment. This prototype chatbot is a new user interface into the GRiST expert-system.
Try Gaia using the web interface. This prototype currently implements the suicide thoughts questions only. Start the interaction using **I feel suicidal**. [Watch the video] [View the chatbot code] [Code for the web UI]
Please have a look at my blog for older projects.
Major League Hacking (MLH) student hackathon at Aston University, November 2015. For a taste of a fun and innovative weekend, please view some videos from the event.