|C V L RAJU|
B .Tech (Mechanical), Nagarjuna UniversityResearch Interests
Project Associate in MEMH Lab, IIT Madras,1994-95Current Work
Reinforcement Learning Algorithms and Applications in Dynamic Pricing of Retail MarketsPublications
My current research focuses on reinforcement learning (RL), and uses electronic commerce as a domain in which to explore the applicability of reinforcement learning algorithms. I am working on an actor-critic simulation based algorithm for the solution of infinite horizon Markov decision process with finite state and action spaces under the discounted cost criterion. The algorithm does gradient search in the space of randomized policies and uses simultaneous perturbation stochastic approximation (SPSA) type estimates. I combine the features of the two-time scale actor-critic algorithms and the gradient search based SPSA technique. I study RL applications in electronic commerce such as, how adaptive software agents may utilize RL algorithms to solve single agent and multi agent problems to make economic decisions such as setting prices in a competitive market place.
C. V. L. Raju and Y. Narahari, Queueing Network Modelling and Lead Time Compression of Electronic Procurement, International Conference on Robotics and Automation (ICRA 2002), Washington D.C., May 11-15, 2002.Contact Information
Shantanu Biswas, C. V. L. Raju, and Y. Narahari, Iterative Reverse Dutch Auction for Procurement of Bundles, Submited to SIAM Conference on Optimization, Toronto, Canada, May 20-22, 2002.
C. V. L. Raju and Y. Narahari, Efficient eProcurement Auctions using Reinforcement Learning, International Conference on Energy, Automation and Information Technology (EAIT 2001), IIT Kharaghpur, December 10-12, 2001.
C. V. L. Raju and Y. Narahari, Effect of Internet Technologies on Performance of Procurement Process, Proceedings of 9th AIMTDR Conference, Indian Institute of Technology Madras, December 2000.
Kaveri # 21,
s/o C. V. N. Raju,