Ness B. Shroff
YOU?
Author Swipe
View article: Online Learning for Optimizing AoI-Energy Tradeoff under Unknown Channel Statistics
Online Learning for Optimizing AoI-Energy Tradeoff under Unknown Channel Statistics Open
We consider a real-time monitoring system where a source node (with energy limitations) aims to keep the information status at a destination node as fresh as possible by scheduling status update transmissions over a set of channels. The fr…
View article: Large Language Models Achieve Gold Medal Performance at the International Olympiad on Astronomy & Astrophysics (IOAA)
Large Language Models Achieve Gold Medal Performance at the International Olympiad on Astronomy & Astrophysics (IOAA) Open
While task-specific demonstrations show early success in applying large language models (LLMs) to automate some astronomical research tasks, they only provide incomplete views of all necessary capabilities in solving astronomy problems, ca…
View article: Discrete Diffusion Models: Novel Analysis and New Sampler Guarantees
Discrete Diffusion Models: Novel Analysis and New Sampler Guarantees Open
Discrete diffusion models have recently gained significant prominence in applications involving natural language and graph data. A key factor influencing their effectiveness is the efficiency of discretized samplers. Among these, $τ$-leapi…
View article: AI Safety vs. AI Security: Demystifying the Distinction and Boundaries
AI Safety vs. AI Security: Demystifying the Distinction and Boundaries Open
Artificial Intelligence (AI) is rapidly being integrated into critical systems across various domains, from healthcare to autonomous vehicles. While its integration brings immense benefits, it also introduces significant risks, including t…
View article: Absorb and Converge: Provable Convergence Guarantee for Absorbing Discrete Diffusion Models
Absorb and Converge: Provable Convergence Guarantee for Absorbing Discrete Diffusion Models Open
Discrete state space diffusion models have shown significant advantages in applications involving discrete data, such as text and image generation. It has also been observed that their performance is highly sensitive to the choice of rate …
View article: Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective
Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective Open
Rehearsal-based methods have shown superior performance in addressing catastrophic forgetting in continual learning (CL) by storing and training on a subset of past data alongside new data in current task. While such a concurrent rehearsal…
View article: Performing Load Balancing under Constraints
Performing Load Balancing under Constraints Open
International audience
View article: Performing Load Balancing under Constraints
Performing Load Balancing under Constraints Open
Join-the-shortest queue (JSQ) and its variants have often been used in solving load balancing problems. The aim of such policies is to minimize the average system occupation, e.g., the customer's system time. In this work we extend the tra…
View article: BeST -- A Novel Source Selection Metric for Transfer Learning
BeST -- A Novel Source Selection Metric for Transfer Learning Open
One of the most fundamental, and yet relatively less explored, goals in transfer learning is the efficient means of selecting top candidates from a large number of previously trained models (optimized for various "source" tasks) that would…
View article: Prediction-Assisted Online Distributed Deep Learning Workload Scheduling in GPU Clusters
Prediction-Assisted Online Distributed Deep Learning Workload Scheduling in GPU Clusters Open
The recent explosive growth of deep learning (DL) models has necessitated a compelling need for efficient job scheduling for distributed deep learning training with mixed parallelisms (DDLwMP) in GPU clusters. This paper proposes an adapti…
View article: How to Find the Exact Pareto Front for Multi-Objective MDPs?
How to Find the Exact Pareto Front for Multi-Objective MDPs? Open
Multi-Objective Markov Decision Processes (MO-MDPs) are receiving increasing attention, as real-world decision-making problems often involve conflicting objectives that cannot be addressed by a single-objective MDP. The Pareto front identi…
View article: Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers
Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers Open
The denoising diffusion model has recently emerged as a powerful generative technique, capable of transforming noise into meaningful data. While theoretical convergence guarantees for diffusion models are well established when the target d…
View article: Can We Theoretically Quantify the Impacts of Local Updates on the Generalization Performance of Federated Learning?
Can We Theoretically Quantify the Impacts of Local Updates on the Generalization Performance of Federated Learning? Open
Federated Learning (FL) has gained significant popularity due to its effectiveness in training machine learning models across diverse sites without requiring direct data sharing. While various algorithms along with their optimization analy…
View article: Artificial Intelligence of Things: A Survey
Artificial Intelligence of Things: A Survey Open
The integration of the Internet of Things (IoT) and modern Artificial Intelligence (AI) has given rise to a new paradigm known as the Artificial Intelligence of Things (AIoT). In this survey, we provide a systematic and comprehensive revie…
View article: Theory on Mixture-of-Experts in Continual Learning
Theory on Mixture-of-Experts in Continual Learning Open
Continual learning (CL) has garnered significant attention because of its ability to adapt to new tasks that arrive over time. Catastrophic forgetting (of old tasks) has been identified as a major issue in CL, as the model adapts to new ta…
View article: Broadening Target Distributions for Accelerated Diffusion Models via a Novel Analysis Approach
Broadening Target Distributions for Accelerated Diffusion Models via a Novel Analysis Approach Open
Accelerated diffusion models hold the potential to significantly enhance the efficiency of standard diffusion processes. Theoretically, these models have been shown to achieve faster convergence rates than the standard $\mathcal O(1/ε^2)$ …
View article: AI‐EDGE: An NSF AI institute for future edge networks and distributed intelligence
AI‐EDGE: An NSF AI institute for future edge networks and distributed intelligence Open
This paper highlights the overall endeavors of the NSF AI Institute for Future Edge Networks and Distributed Intelligence (AI‐EDGE) to create a research, education, knowledge transfer, and workforce development environment for developing t…
View article: Model-Free Change Point Detection for Mixing Processes
Model-Free Change Point Detection for Mixing Processes Open
This paper considers the change point detection problem under dependent samples. In particular, we provide performance guarantees for the MMD-CUSUM test under exponentially , , and fast -mixing processes, which significantly expands its ut…
View article: Model-Free Change Point Detection for Mixing Processes
Model-Free Change Point Detection for Mixing Processes Open
This paper considers the change point detection problem under dependent samples. In particular, we provide performance guarantees for the MMD-CUSUM test under exponentially $α$, $β$, and fast $ϕ$-mixing processes, which significantly expan…
View article: Optimal Edge Caching For Individualized Demand Dynamics
Optimal Edge Caching For Individualized Demand Dynamics Open
The ever-growing end user data demands, and the simultaneous reductions in memory costs are fueling edge-caching deployments. Caching at the edge is substantially different from that at the core and needs to take into account the nature of…
View article: Sampling for Remote Estimation of the Wiener Process over an Unreliable Channel
Sampling for Remote Estimation of the Wiener Process over an Unreliable Channel Open
In this paper, we study a sampling problem where a source takes samples from a Wiener process and transmits them through a wireless channel to a remote estimator. Due to channel fading, interference, and potential collisions, the packet tr…
View article: Hoeffding's Inequality for Markov Chains under Generalized Concentrability Condition
Hoeffding's Inequality for Markov Chains under Generalized Concentrability Condition Open
This paper studies Hoeffding's inequality for Markov chains under the generalized concentrability condition defined via integral probability metric (IPM). The generalized concentrability condition establishes a framework that interpolates …
View article: Near Delay-Optimal Scheduling of Batch Jobs in Multi-Server Systems
Near Delay-Optimal Scheduling of Batch Jobs in Multi-Server Systems Open
We study a class of scheduling problems, where each job is divided into a batch of unit-size tasks and these tasks can be executed in parallel on multiple servers with New-Better-than-Used (NBU) service time distributions. While many delay…
View article: Energy-Efficient Deadline-Aware Edge Computing: Bandit Learning with Partial Observations in Multi-Channel Systems
Energy-Efficient Deadline-Aware Edge Computing: Bandit Learning with Partial Observations in Multi-Channel Systems Open
In this paper, we consider a task offloading problem in a multi-access edge computing (MEC) network, in which edge users can either use their local processing unit to compute their tasks or offload their tasks to a nearby edge server throu…
View article: Non-Convex Bilevel Optimization with Time-Varying Objective Functions
Non-Convex Bilevel Optimization with Time-Varying Objective Functions Open
Bilevel optimization has become a powerful tool in a wide variety of machine learning problems. However, the current nonconvex bilevel optimization considers an offline dataset and static functions, which may not work well in emerging onli…
View article: Achieving Sample and Computational Efficient Reinforcement Learning by Action Space Reduction via Grouping
Achieving Sample and Computational Efficient Reinforcement Learning by Action Space Reduction via Grouping Open
Reinforcement learning often needs to deal with the exponential growth of states and actions when exploring optimal control in high-dimensional spaces (often known as the curse of dimensionality). In this work, we address this issue by lea…
View article: Theoretical Hardness and Tractability of POMDPs in RL with Partial Online State Information
Theoretical Hardness and Tractability of POMDPs in RL with Partial Online State Information Open
Partially observable Markov decision processes (POMDPs) have been widely applied in various real-world applications. However, existing theoretical results have shown that learning in POMDPs is intractable in the worst case, where the main …