Learn to Solve Constrained Markov Decision Process Efficiently

#WIE #engineering #photonics #science
Share

Learn to Solve Constrained Markov Decision Process Efficiently

 

Abstract: Many constrained sequential decision-making processes, such as safe AV navigation, wireless network control, caching, cloud computing, etc., can be cast as Constrained Markov Decision Processes (CMDP). Reinforcement Learning (RL) algorithms have been used to learn optimal policies for unknown unconstrained MDP. Extending these RL algorithms to unknown CMDP brings the additional challenge of maximizing the reward and satisfying the constraints. In this talk, I will present algorithms that can learn safe policies effectively. 

In the second part of the talk, I will demonstrate how the theoretical understanding of the constrained MDP can help us to develop algorithms for practical applications. As an application, I show how to learn to obtain optimal beam directions under time-varying interference-constrained channels for a mobile service robot. Optimal beam selection in mmWave is challenging because of its time-varying nature. We propose a primal-dual Gaussian process bandit with adaptive reinitialization to handle non-stationarity and interference constraints. We demonstrate how our approach learns to adapt effectively to time-varying channel conditions.

 


  Date and Time

  Location

  Hosts

  Registration


  • Start time: 06 Nov 2024 07:00 PM
  • End time: 07 Nov 2024 08:00 PM
  • All times are (UTC-05:00) Eastern Time (US & Canada)
  • Add_To_Calendar_icon Add Event to Calendar
  • 154 Summit Street
  • Newark, New Jersey
  • United States 07102
  • Building: Electrical and Computer Engineering
  • Room Number: ECE 202
  • Contact Event Hosts
  • Co-sponsored by IEEE North Jersey Section
  • Starts 22 October 2024 12:00 AM
  • Ends 08 November 2024 12:00 AM
  • All times are (UTC-05:00) Eastern Time (US & Canada)
  • No Admission Charge

  Speakers

Dr. Arnob Ghosh

Topic:

Learn to Solve Constrained Markov Decision Process Efficiently

Arnob Ghosh

Assistant Professor,

Electrical and Computer Engineering Department,

New Jersey Institute of Technology,

Office Phone: 973-596-3342

 

 

 

Biography:

Arnob Ghosh has been an Assistant Professor at the Department of Electrical and Computer Engineering at New Jersey Institute of Technology (NJIT) since Fall 2023.
Before joining NJIT, he was a Research Scientist at Ohio State University's Electrical and Computer Engineering Department. 
He was also part of the NSF AI-EDGE Institute (https://aiedge.osu.edu/). Ness Shroff has hosted him since June 2021. Before that, he was an Assistant Professor at the Industrial Engineering and Operations Research (IEOR) Program of the Mechanical Engineering Department at the Indian Institute of Technology-Delhi (IIT-Delhi) since August 2019. At IIT-Delhi, he was also associated with the School of Artificial Intelligence. Before joining IIT-Delhi, he was a post-doctoral research associate at the School of Industrial Engineering at Purdue University from August 2016 to July 2019. He was hosted by Dr. Vaneet Aggarwal there.

He completed his Ph.D. in August 2016 from the Electrical and System Engineering Department at the University of Pennsylvania, where his advisor was Prof. Saswati Sarkar. Before joining UPENN, he studied Electronics & Telecommunication Engineering at Jadavpur University from 2007-2011.
 
Education
Ph.D.; University of Pennsylvania; Electrical and Systems Engineering; 2016
M.S.; University of Pennsylvania; Electrical Engineering; 2013

B.E.; Jadavpur University; Electronics and Tele-Communication Engineering; 2011

Website
https://sites.google.com/site/arnob008/
 
 

Address:United States




Agenda

Nov 6th,

Talk: 7:00 PM - 8:00 PM

Discussion Q/A: 8:00 PM - 8:15 PM