PhD Dissertation Defense by Meghna LOWALEKAR | Online Spatio-Temporal Demand Supply Matching

About The Talk

The rapid growth of cities in developing world coupled with the increase in rural to urban migration, led to cities being identified as the key actor for any nation’s economy. Shared mobility has transformed many global cities by improving efficiency and enhancing transportation accessibility. The fast paced innovations in internet technologies have enabled shared mobility to become an integral part of life of people in cities. The mismatch between the demand and supply of these shared mobility resources has a direct impact on people’s life as apart from causing inconvenience to people, it can also contribute towards increasing pollution and congestion on the road by leading to extensive usage of private vehicles.. Thus in my dissertation, I focus on developing solution strategies for these real-time (online) spatio-temporal demand supply matching problems which can enhance the service quality of shared mobility resources by considering expected future demand. To make better sequential and connected decisions, I use data driven multi-period two-stage optimization approaches by taking multiple samples of customer requests from the historical data. In the proposed algorithms, to handle the large scale nature of problems, I exploit the homogeneous nature of resources by using abstraction based techniques. The scalability is further improved by using decomposition based approaches which break a large problem into multiple smaller problems which can be solved in parallel. Experimental evaluation on real-world datasets show that the proposed approaches significantly outperform existing approaches.

Speaker Biography

Meghna Lowalekar is a Ph.D. candidate in School of Information Systems, Singapore Management University. She is working under the supervision of Associate Professor Pradeep Varakantham from SMU and Professor Patrick Jaillet from MIT. She received her B.Tech. & M.S in Computer Science & Engineering from IIIT, Hyderabad, India. Prior to joining the PhD program, she has also worked as a software engineer in Qualcomm and Microsoft in Hyderabad, India. She has published papers at prestigious AI journals (JAIR, AIJ) and conferences (AAAI, AAMAS, ICAPS). Her work received Best Application Paper Award at ICAPS 2019 and Best Demo Award at AAMAS 2018.