COMPUTER SCIENCE & ENGINEERING
Dr. Ajin George Joseph
Areas of Interest
- Reinforcement learning
- Stochastic approximation
- Stochastic search
- Simulation optimization
- Markov decision processes
- Deep learning
- B.Tech, National Institute of Technology, Calicut, India
- M.Tech, Indian Institute of Technology, Kanpur, India
- Ph.D, Indian Institute of Science, Bangalore, India
- A.G. Joseph and S. Bhatnagar, An Adaptive and incremental approach
to quantile estimation, IEEE Conference on Decision and Control
- Two timescale networks for nonlinear value function approximation ,
Wesley Chung, Somjit Nath, A.G. Joseph and Martha White, ICLR,
- Stochastic approximation trackers for model based search method,
A.G. Joseph and S. Bhatnagar, Allerton Conference on Communication, Control, and Computing, 2019.
- An online prediction algorithm for reinforcement learning with linear
function approximation using cross entropy method, A.G. Joseph and
S. Bhatnagar, Machine Learning Journal, 2018.
An incremental off policy search in a model free Markov decision process using a single sample path, A.G. Joseph and S. Bhatnagar, Machine Learning Journal, 2018.