Akshay Mete

Akshay Mete

PhD Student in Reinforcement Learning at Texas A&M University, College Station, TX, USA

View Research Get in Touch

About Me

I am a PhD student at Texas A&M University, specializing in reinforcement learning, adaptive control, and applied machine learning.

Research Projects

The Exploration-Exploitation Trade-Off in Reinforcement Learning

Reward Biased Maximum Likelihood Estimation (RBMLE) algorithm, an online model-based RL algorithm

  • Addresses excessive exploration in UCB algorithms
  • Incorporates controlled exploration without explicit confidence intervals
  • Demonstrates promising performance in MDPs, linear time series, safe RL, multi-objective RL, and RLHF for Language Models

Reinforcement Learning for Auto-Regressive Time Series

PIECE, a self-tuning regulator for linear time series with \( \mathcal{O}(\log T) \) regret

  • First \( \mathcal{O}(\log T) \) regret bound for self-tuning regulation
  • Outperforms traditional control algorithms

Publications

Reinforcement Learning

Augmented RBMLE-UCB Approach for Adaptive Control of Linear Quadratic Systems

Advances in Neural Information Processing Systems (NeurIPS) 2022

A. Mete, R. Singh, and P. R. Kumar

Paper

Reward Biased Maximum Likelihood Estimation for Reinforcement Learning

Learning for Dynamics and Control (L4DC) 2021

A. Mete, R. Singh, Xi Liu, and P. R. Kumar

Paper

Finite Time Regret Bounds for Self-Tuning Regulation

International Conference on Machine Learning (ICML) 2024

R. Singh, A. Mete, A. Kar, and P. R. Kumar

Paper

The Reward Biased Method: An Optimism based Approach for Reinforcement Learning

Allerton Conference on Communication, Control, and Computing 2023 (Invited Paper)

A. Mete, R. Singh, and P. R. Kumar

Paper

The RBMLE Method for Reinforcement Learning

Conference on Information Sciences and Systems (CISS) 2022 (Invited Paper)

A. Mete, R. Singh, and P. R. Kumar

Paper

Game Theory for Networks

Partial Server Pooling in Redundancy Systems

International Teletraffic Congress (ITC 31) 2019

A. Mete, D. Manjunath, J. Nair, and B. Prabhu

Paper

Partial Server Pooling in Delay Systems

ACM Transactions on Modeling and Performance Evaluation of Computing Systems 2024

V. Rohit, A. Mete, D. Manjunath, J. Nair, and B. Prabhu

Paper

Caching Policies for D2D-assisted Content Delivery Systems

Technologies for the Wireless Edge Workshop, MobiCom 2018

A. Mete and S. Moharir

Paper

Low Latency Scheduling for D2D Communication

International Conference on Communication Systems & Networks (COMSNETS) 2019

A. Mete

Paper

CyberSecurity

Cyber-Secure and Safe Operation of Solar Photovoltaic Power Distribution Systems

IEEE Applied Power Electronics Conference and Exposition (APEC) 2024

J. Kim, H. Ibrahim, S. Wang, A. Mete, L. Xie, P. Enjeti, and P.R. Kumar

Paper

Advisors

P. R. Kumar

Texas A&M University

PhD Advisor

D. Manjunath

IIT Bombay

MS Advisor

R. Singh

IISc Bengaluru

Research Collaborator

J. Nair

IIT Bombay

MS Advisor

S. Moharir

IIT Bombay

MS Advisor

H. Wang

AT&T Labs

Research Internship Mentor

Professional Experience

Research Intern

Network Analytics and Automation Group, AT&T Labs, Bedminster, NJ, USA

Summer 2024

Reinforcement Learning for Network Slicing

Education

PhD, Electrical and Computer Engineering

Texas A&M University, College Station, TX, USA

2019 - Present

Advisor: Prof. P. R. Kumar

Thesis: The Exploration-Exploitation Trade-Off in Reinforcement Learning

MS + BS (Dual Degree), Electrical Engineering

Indian Institute of Technology (IIT) Bombay, Mumbai, India

2013 - 2018

Specialization: Communication and Signal Processing

Relevant Courses

Artificial Intelligence

  • Reinforcement Learning
  • Machine Learning
  • Deep Learning

Optimization

  • Convex Optimization
  • Combinatorial Optimization
  • Stochastic Optimization

Control Theory

  • Stochastic Systems
  • Adaptive Control
  • Linear Control Systems

Mathematics and Statistics

  • Real Analysis
  • Stochastic Processes
  • Graph Theory

Academic Activities

Conference Reviewer

  • International Conference on Machine Learning (ICML)
  • Neural Information Processing Systems (NeurIPS)
  • International Conference on Artificial Intelligence and Statistics (AISTATS)
  • Association for the Advancement of Artificial Intelligence (AAAI)
  • Learning for Dynamics and Control (L4DC)
  • IEEE Conference on Decision and Control (CDC)
  • American Control Conference (ACC)
  • Indian Control Conference (ICC)

Seminars & Tutorials

  • "Reinforcement Learning: A Control Theoretic Perspective", Long Tutorial, ICC 2023
  • "From Adaptive Control to Reinforcement Learning", Purdue University

Get in Touch

lastnamefirstname@gmail.com

College Station, TX, USA

© 2025 Akshay Mete. All rights reserved.