Search Results for author: Dileep Kalathil

Found 39 papers, 14 papers with code

Structured Reinforcement Learning for Media Streaming at the Wireless Edge

no code implementations10 Apr 2024 Archana Bura, Sarat Chandra Bobbili, Shreyas Rameshkumar, Desik Rengarajan, Dileep Kalathil, Srinivas Shakkottai

The goal of this work is to develop and demonstrate learning-based policies for optimal decision making to determine which clients to dynamically prioritize in a video streaming setting.

reinforcement-learning

AttackGNN: Red-Teaming GNNs in Hardware Security Using Reinforcement Learning

no code implementations21 Feb 2024 Vasudev Gohil, Satwik Patnaik, Dileep Kalathil, Jeyavijayan Rajendran

We target five GNN-based techniques for four crucial classes of problems in hardware security: IP piracy, detecting/localizing HTs, reverse engineering, and hardware obfuscation.

Graph Neural Network reinforcement-learning +1

Meta-Learning-Based Adaptive Stability Certificates for Dynamical Systems

1 code implementation23 Dec 2023 Amit Jena, Dileep Kalathil, Le Xie

This paper addresses the problem of Neural Network (NN) based adaptive stability certification in a dynamical system.

Meta-Learning

Bridging Distributionally Robust Learning and Offline RL: An Approach to Mitigate Distribution Shift and Partial Data Coverage

1 code implementation27 Oct 2023 Kishan Panaganti, Zaiyan Xu, Dileep Kalathil, Mohammad Ghavamzadeh

The goal of an offline reinforcement learning (RL) algorithm is to learn optimal polices using historical (offline) data, without access to the environment for online exploration.

Offline RL Reinforcement Learning (RL)

Natural Actor-Critic for Robust Reinforcement Learning with Function Approximation

1 code implementation NeurIPS 2023 Ruida Zhou, Tao Liu, Min Cheng, Dileep Kalathil, P. R. Kumar, Chao Tian

We study robust reinforcement learning (RL) with the goal of determining a well-performing policy that is robust against model mismatch between the training simulator and the testing environment.

reinforcement-learning Reinforcement Learning (RL)

LLMZip: Lossless Text Compression using Large Language Models

2 code implementations6 Jun 2023 Chandra Shekhara Kaushik Valmeekam, Krishna Narayanan, Dileep Kalathil, Jean-Francois Chamberland, Srinivas Shakkottai

We provide new estimates of an asymptotic upper bound on the entropy of English using the large language model LLaMA-7B as a predictor for the next token given a window of past tokens.

Language Modelling Large Language Model +1

Federated Ensemble-Directed Offline Reinforcement Learning

1 code implementation4 May 2023 Desik Rengarajan, Nitin Ragothaman, Dileep Kalathil, Srinivas Shakkottai

We consider the problem of federated offline reinforcement learning (RL), a scenario under which distributed learning agents must collaboratively learn a high-quality control policy only using small pre-collected datasets generated according to different unknown behavior policies.

Continuous Control Ensemble Learning +4

Improved Sample Complexity Bounds for Distributionally Robust Reinforcement Learning

1 code implementation5 Mar 2023 Zaiyan Xu, Kishan Panaganti, Dileep Kalathil

We formulate this as a distributionally robust reinforcement learning (DR-RL) problem where the objective is to learn the policy which maximizes the value function against the worst possible stochastic model of the environment in an uncertainty set.

reinforcement-learning Reinforcement Learning (RL)

Dynamic Regret Analysis of Safe Distributed Online Optimization for Convex and Non-convex Problems

no code implementations23 Feb 2023 Ting-Jui Chang, Sapana Chaudhary, Dileep Kalathil, Shahin Shahrampour

We prove that for convex functions, D-Safe-OGD achieves a dynamic regret bound of $O(T^{2/3} \sqrt{\log T} + T^{1/3}C_T^*)$, where $C_T^*$ denotes the path-length of the best minimizer sequence.

Enhanced Meta Reinforcement Learning using Demonstrations in Sparse Reward Environments

1 code implementation26 Sep 2022 Desik Rengarajan, Sapana Chaudhary, Jaewon Kim, Dileep Kalathil, Srinivas Shakkottai

Meta reinforcement learning (Meta-RL) is an approach wherein the experience gained from solving a variety of tasks is distilled into a meta-policy.

Meta Reinforcement Learning reinforcement-learning +1

Meta-Learning Online Control for Linear Dynamical Systems

no code implementations18 Aug 2022 Deepan Muthirayan, Dileep Kalathil, Pramod P. Khargonekar

We show that when the number of tasks are sufficiently large, our proposed approach achieves a meta-regret that is smaller by a factor $D/D^{*}$ compared to an independent-learning online control algorithm which does not perform learning across the tasks, where $D$ is a problem constant and $D^{*}$ is a scalar that decreases with increase in the similarity between tasks.

Meta-Learning

Robust Reinforcement Learning using Offline Data

1 code implementation10 Aug 2022 Kishan Panaganti, Zaiyan Xu, Dileep Kalathil, Mohammad Ghavamzadeh

The goal of robust reinforcement learning (RL) is to learn a policy that is robust against the uncertainty in model parameters.

reinforcement-learning Reinforcement Learning (RL)

Distributed Learning of Neural Lyapunov Functions for Large-Scale Networked Dissipative Systems

no code implementations15 Jul 2022 Amit Jena, Tong Huang, S. Sivaranjani, Dileep Kalathil, Le Xie

One standard approach to estimate the stability region of a general nonlinear system is to first find a Lyapunov function for the system and characterize its region of attraction as the stability region.

Distributed Optimization

Anchor-Changing Regularized Natural Policy Gradient for Multi-Objective Reinforcement Learning

2 code implementations10 Jun 2022 Ruida Zhou, Tao Liu, Dileep Kalathil, P. R. Kumar, Chao Tian

We study policy optimization for Markov decision processes (MDPs) with multiple reward value functions, which are to be jointly optimized according to given criteria such as proportional fairness (smooth concave scalarization), hard constraints (constrained MDP), and max-min trade-off.

Fairness Multi-Objective Reinforcement Learning +1

The Impact of Heavy-Duty Vehicle Electrification on Large Power Grids: a Synthetic Texas Case Study

no code implementations8 Mar 2022 Rayan El Helou, S. Sivaranjani, Dileep Kalathil, Andrew Schaper, Le Xie

In fact, we find that as little as 11% of heavy duty vehicles in Texas charging simultaneously can lead to significant voltage violations on the transmission network that compromise grid reliability.

Reinforcement Learning with Sparse Rewards using Guidance from Offline Demonstration

1 code implementation ICLR 2022 Desik Rengarajan, Gargi Vaidya, Akshay Sarvesh, Dileep Kalathil, Srinivas Shakkottai

We demonstrate the superior performance of our algorithm over state-of-the-art approaches on a number of benchmark environments with sparse rewards and censored state.

reinforcement-learning Reinforcement Learning (RL)

Off-Policy Evaluation Using Information Borrowing and Context-Based Switching

1 code implementation18 Dec 2021 Sutanoy Dasgupta, Yabo Niu, Kishan Panaganti, Dileep Kalathil, Debdeep Pati, Bani Mallick

We consider the off-policy evaluation (OPE) problem in contextual bandits, where the goal is to estimate the value of a target policy using the data collected by a logging policy.

Multi-Armed Bandits Off-policy evaluation

Sample Complexity of Robust Reinforcement Learning with a Generative Model

1 code implementation2 Dec 2021 Kishan Panaganti, Dileep Kalathil

For each of these uncertainty sets, we give a precise characterization of the sample complexity of our proposed algorithm.

Model-based Reinforcement Learning reinforcement-learning +1

DOPE: Doubly Optimistic and Pessimistic Exploration for Safe Reinforcement Learning

1 code implementation1 Dec 2021 Archana Bura, Aria HasanzadeZonuzy, Dileep Kalathil, Srinivas Shakkottai, Jean-Francois Chamberland

Safe reinforcement learning is extremely challenging--not only must the agent explore an unknown environment, it must do so while ensuring no safety constraint violations.

reinforcement-learning Reinforcement Learning (RL) +2

Online Learning for Predictive Control with Provable Regret Guarantees

no code implementations30 Nov 2021 Deepan Muthirayan, Jianjun Yuan, Dileep Kalathil, Pramod P. Khargonekar

Specifically, we study the online learning problem where the control algorithm does not know the true system model and has only access to a fixed-length (that does not grow with the control horizon) preview of the future cost functions.

Model Predictive Control

Safe Online Convex Optimization with Unknown Linear Safety Constraints

no code implementations14 Nov 2021 Sapana Chaudhary, Dileep Kalathil

We study the problem of safe online convex optimization, where the action at each time step must satisfy a set of linear safety constraints.

Policy Optimization for Constrained MDPs with Provable Fast Global Convergence

no code implementations31 Oct 2021 Tao Liu, Ruida Zhou, Dileep Kalathil, P. R. Kumar, Chao Tian

We propose a new algorithm called policy mirror descent-primal dual (PMD-PD) algorithm that can provably achieve a faster $\mathcal{O}(\log(T)/T)$ convergence rate for both the optimality gap and the constraint violation.

OTTR: Off-Road Trajectory Tracking using Reinforcement Learning

no code implementations5 Oct 2021 Akhil Nagariya, Dileep Kalathil, Srikanth Saripalli

Compared to the standard ILQR approach, our proposed approach achieves a 30% and 50% reduction in cross track error in Warthog and Moose, respectively, by utilizing only 30 minutes of real-world driving data.

reinforcement-learning Reinforcement Learning (RL)

PyProD: A Machine Learning-Friendly Platform for Protection Analytics in Distribution Systems

no code implementations13 Sep 2021 Dongqi Wu, Dileep Kalathil, Miroslav Begovic, Le Xie

This paper introduces PyProD, a Python-based machine learning (ML)-compatible test-bed for evaluating the efficacy of protection schemes in electric distribution grids.

BIG-bench Machine Learning Decision Making

Learning Policies with Zero or Bounded Constraint Violation for Constrained MDPs

no code implementations NeurIPS 2021 Tao Liu, Ruida Zhou, Dileep Kalathil, P. R. Kumar, Chao Tian

We show that when a strictly safe policy is known, then one can confine the system to zero constraint violation with arbitrarily high probability while keeping the reward regret of order $\tilde{\mathcal{O}}(\sqrt{K})$.

Safe Exploration

Fully Decentralized Reinforcement Learning-based Control of Photovoltaics in Distribution Grids for Joint Provision of Real and Reactive Power

no code implementations3 Aug 2020 Rayan El Helou, Dileep Kalathil, Le Xie

In this paper, we introduce a new framework to address the problem of voltage regulation in unbalanced distribution grids with deep photovoltaic penetration.

reinforcement-learning Reinforcement Learning (RL)

Learning with Safety Constraints: Sample Complexity of Reinforcement Learning for Constrained MDPs

no code implementations1 Aug 2020 Aria HasanzadeZonuzy, Archana Bura, Dileep Kalathil, Srinivas Shakkottai

Many physical systems have underlying safety considerations that require that the policy employed ensures the satisfaction of a set of constraints.

reinforcement-learning Reinforcement Learning (RL)

Reinforcement Learning for Mean Field Games with Strategic Complementarities

no code implementations21 Jun 2020 Kiyeob Lee, Desik Rengarajan, Dileep Kalathil, Srinivas Shakkottai

We introduce a natural refinement to the equilibrium concept that we call Trembling-Hand-Perfect MFE (T-MFE), which allows agents to employ a measure of randomization while accounting for the impact of such randomization on their payoffs.

reinforcement-learning Reinforcement Learning (RL)

Robust Reinforcement Learning using Least Squares Policy Iteration with Provable Performance Guarantees

no code implementations20 Jun 2020 Kishan Panaganti, Dileep Kalathil

We first propose the Robust Least Squares Policy Evaluation algorithm, which is a multi-step online model-free learning algorithm for policy evaluation.

OpenAI Gym reinforcement-learning +1

Deep Reinforcement Learning-BasedRobust Protection in DER-Rich Distribution Grids

no code implementations5 Mar 2020 Dongqi Wu, Dileep Kalathil, Miroslav Begovic, Le Xie

This paper introduces the concept of Deep Reinforcement Learning based architecture for protective relay design in power distribution systems with many distributed energy resources (DERs).

reinforcement-learning Reinforcement Learning (RL)

Bounded Regret for Finitely Parameterized Multi-Armed Bandits

no code implementations3 Mar 2020 Kishan Panaganti, Dileep Kalathil

We propose an algorithm that is simple and easy to implement, which we call Finitely Parameterized Upper Confidence Bound (FP-UCB) algorithm, which uses the information about the underlying parameter set for faster learning.

Multi-Armed Bandits

Decoupled Data Based Approach for Learning to Control Nonlinear Dynamical Systems

1 code implementation17 Apr 2019 Ran Wang, Karthikeya Parunandi, Dan Yu, Dileep Kalathil, Suman Chakravorty

This paper proposes a novel decoupled data-based control (D2C) algorithm that addresses this problem using a decoupled, `open loop - closed loop', approach.

QFlow: A Learning Approach to High QoE Video Streaming at the Wireless Edge

no code implementations4 Jan 2019 Rajarshi Bhattacharyya, Archana Bura, Desik Rengarajan, Mason Rumuly, Bainan Xia, Srinivas Shakkottai, Dileep Kalathil, Ricky K. P. Mok, Amogh Dhamdhere

The predominant use of wireless access networks is for media streaming applications, which are only gaining popularity as ever more devices become available for this purpose.

On Regret-Optimal Learning in Decentralized Multi-player Multi-armed Bandits

no code implementations4 May 2015 Naumaan Nayyar, Dileep Kalathil, Rahul Jain

The objective is to design a policy that maximizes the expected reward over a time horizon for a single player setting and the sum of expected rewards for the multiplayer setting.

Multi-Armed Bandits

Empirical Q-Value Iteration

no code implementations30 Nov 2014 Dileep Kalathil, Vivek S. Borkar, Rahul Jain

We propose a new simple and natural algorithm for learning the optimal Q-value function of a discounted-cost Markov Decision Process (MDP) when the transition kernels are unknown.

Q-Learning

Approachability in Stackelberg Stochastic Games with Vector Costs

no code implementations3 Nov 2014 Dileep Kalathil, Vivek Borkar, Rahul Jain

Firstly, we give a simple and computationally tractable strategy for approachability for Stackelberg stochastic games along the lines of Blackwell's.

Decision Making

Cannot find the paper you are looking for? You can Submit a new open access paper.