Online Learning

Follow this topic Ask about this topic

Online learning studies how to make predictions and decisions as data arrive, rather than training once on a fixed dataset. Common approaches include Online Mirror Descent and Follow-the-Regularized-Leader for optimization, exponential-weights methods for choosing among experts, and bandit algorithms when feedback is limited. Researchers analyze regret and mistake bounds in adversarial, stochastic, and changing environments, and apply these methods to areas such as forecasting, auctions, and resource allocation. A central challenge is adapting quickly to change while remaining stable and efficient when feedback is delayed, noisy, or costly.

Latest papers

132 papers24 papers in the last four weeks, up 60% on the four weeks before. 0.1% of all new papers.

Gradient-Free Warm-Start Library Recovery: an Amortized-Regret Separation
Jun 19, 2026Jianwei LouGradient-Free Optimization Continual Learning
RailMind Systems, Neuss, Germany
Provably Sub-Linear Two-Timescale NeuroEvolution with Online Plasticity
Jun 18, 2026Shishen Lin, Yixin ChenRL Control Evolutionary Optimization
Department of Statistics, University of Warwick, Coventry, United Kingdom · 2Biozentrum, University of Basel, Basel, Switzerland
Online Convex Optimization with Sublinear Noisy Probes
Jun 12, 2026Simone Di Gregorio, Anupam Gupta, Stefano Leonardi +1Online Convex Optimization Online Learning
Sapienza University of Rome, Italy · New York University, USA · EPFL, Switzerland
A Programmer's Guide to Cascaded Adaptive Combiners: Online Learning by Biologically Accurate Models of Multilayer Neuron Networks
Jun 12, 2026Martin Nilsson, Denis KleykoBiologically Plausible Learning Adaptive Learning
RISE Research Institutes of Sweden, Kista, Sweden · Örebro University, Örebro, Sweden
Structured Adaptive Tensor Prediction for Streaming Data
Jun 8, 2026Zhen Qin, Yang ChenTime Series Forecasting Online Learning
Michigan Institute for Computational Discovery and Engineering, Department of Statistics, Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI 48109 USA · Department of Statistics, University of Michigan, Ann Arbor, MI 48109 USA
A Switching Beamformer for Highly Non-Stationary Environments
Jun 7, 2026Manan Mittal, Ryan M. Corey, John R. Buck +1Online Learning
Electrical and Computer Engineering, Stony Brook University, Stony Brook, NY 11794, USA · Electrical and Computer Engineering, University of Illinois Chicago, Chicago, IL 60607, USA · Electrical and Computer Engineering, University of Massachusetts Dartmouth, Dartmouth, MA 02747, USA +1
AdaKoop: Efficient Modeling of Nonlinear Dynamics from Nonstationary Data Streams with Koopman Operator Regression
Jun 3, 2026Naoki Chihara, Ren Fujiwara, Yasuko Matsubara +1Koopman Operator Learning Nonlinear System Identification
SANKEN, The University of Osaka Osaka, Japan
An RRAM-based Hardware Implementation of a Radial Basis Function Neuron for Edge Classifiers
Jun 2, 2026Georgios Papandroulidakis, Shady Agwa, Themis ProdromakisCompute-in-Memory Edge ML
Centre for Electronics Frontiers, Institute of Micro and Nano Systems, School of Engineering, The University of Edinburgh, EH9 3BF, UK.
AUGUSTE: Online-Learning dApp for Predictive URLLC Scheduling
Jun 2, 2026Maxime Elkael, Michele Polese, Yunseong Lee +2Wireless Communications Wireless Resource Allocation
Institute for Intelligent Networked Systems, Northeastern University (NU), Boston, MA, USA. · SoftBank Corp. (SB), Japan.
AdaWeather: Adaptively Mixing Probabilistic Weather Forecasts with Logarithmic Regret
Jun 1, 2026Saptarishi Dhanuka, Sarvesh Iyer, Manmeet Singh +5Probabilistic Forecasting Mixture-of-Experts Models
Ashoka University · Ashoka University, India · Western Kentucky University
Lodestar: An Online-Learning LLM Inference Router
May 31, 2026Gangmuk Lim, Wanyu Zhao, Brighten Godfrey +3Inference-Time Optimization LLM Inference Scheduling
UIUC · Bytedance · University of Edinburgh
Optimal Gap-Dependent Regret for Private Stochastic Decision-Theoretic Online Learning
May 27, 2026Tommaso Cesari, Roberto ColomboniDifferential Privacy Online Learning
School of Electrical Engineering and Computer Science University of Ottawa Ottawa, Canada · School of Mathematics University of Bristol Bristol, United Kingdom
The Fundamental Limits of Fraud Detection in Card Payment Networks
May 26, 2026Gaurav DhamaFinancial Fraud Detection Online Learning
Mastercard
Online Learning on Hidden-Convex Losses via Algorithmic Equivalence: Optimal Regret, Geometric Barrier, and Bandit Feedback
May 25, 2026Anas Barakat, Andreas Kontogiannis, Vasilis Pollatos +2Online Convex Optimization Gradient Descent
Singapore University of Technology and Design · National Technical University of Athens · Archimedes, Athena Research Center, Greece +3
A Boundary-Layer Mechanism for One-Third Scaling in Online Softmax Classification
May 21, 2026Marcel Kühn, Yoon Thelge, Bernd RosenowNeural Network Training Dynamics Multiclass Classification
Institute for Theoretical Physics, Leipzig University · ScaDS.AI Dresden/Leipzig
A Geometric Approach to Constrained Online Learning
May 20, 2026Dhruv Sarkar, Abhishek SinhaOnline Convex Optimization Online Learning
Indian Institute of Technology Kharagpur · Mohamed bin Zayed University of Artificial Intelligence · Tata Institute of Fundamental Research, Mumbai
Factor Augmented High-Dimensional SGD
May 19, 2026Shubo Li, Yuefeng Han, Xiufan YuStochastic Optimization Stochastic Optimization Convergence
Department of Statistics The Pennsylvania State University University Park, PA 16802, USA · Department of Applied and Computational Mathematics and Statistics University of Notre Dame Notre Dame, IN 46556, USA
Self-Regulated Learning in Essay Writing: Consistency of Strategies and Impact on Outcomes
May 14, 2026Gloria Fernández-Nieto, Kiyoshige Garcés, Mladen Raković +4Self-Regulated Learning Online Learning
Department of Data Science & AI Monash University Australia, Victoria
Always Learning, Always Mixing: Efficient and Simple Data Mixing All The Time
May 13, 2026Michael Y. Hu, Apurva Gandhi, Kyunghyun Cho +2Language Model Pretraining Continual Learning for LLMs
New York University · Carnegie Mellon University · Microsoft
Learning to Decide with AI Assistance under Human-Alignment
May 12, 2026Nina Corvelo Benz, Eleni Straitouri, Manuel Gomez-RodriguezAI-Assisted Decision Making Human-AI Decision Making
§Max Planck Institute for Biochemistry, Martinsried, Germany · †Max Planck Institute for Software Systems, Kaiserslautern, Germany
Online Learning-to-Defer with Varying Experts
May 12, 2026Dang Hoang Duy, Yannis Montreuil, Maxime Meyer +3Learning to Defer Bandit Learning
School of Computing, National University of Singapore, Singapore · IPAL, IRL 2955, Singapore · Department of Mathematics, National University of Singapore, Singapore, 117543 +2
Continuous Latent Contexts Enable Efficient Online Learning in Transformers
May 11, 2026Emile Anand, Abdullah Ateyeh, Xinyuan Cao +1Transformer Sequential Decision Making
Georgia Institute of Technology · University of California, Berkeley · Columbia University
Online Set Learning from Precision and Recall Feedback
May 10, 2026Lee Cohen, Yishay Mansour, Shay Moran +1PAC Learning Statistical Learning Theory
1Stanford University · 2Tel Aviv University and Google Research · 3Technion and Google Research +1
Learnability and Competition in High-Dimensional Multi-Component ICA
May 8, 2026Eser Ilke Genc, Samet Demir, Zafer DoganIndependent Component Analysis Unsupervised Representation Learning
MLIP Research Group, KUIS AI Center, Koç University · Department of EEE, Koç University
Self-Play Enhancement via Advantage-Weighted Refinement in Online Federated LLM Fine-Tuning
May 8, 2026Seohyun Lee, Wenzhi Fang, Dong-Jun Han +2LLM Fine-Tuning Language Model Self-Improvement
Purdue University · Yonsei University · University at Buffalo-SUNY
Characterizing and Correcting Effective Target Shift in Online Learning
May 8, 2026Ziyan Li, Naoki HirataniLabel Shift Kernel Regression
Department of Physics Washington University in St. Louis St. Louis, USA · Department of Neuroscience Washington University in St. Louis St. Louis, USA
Regret-Oracle Complexity Tradeoffs in Agnostic Online Learning
May 8, 2026Idan Attias, Steve Hanneke, Arvind RamaswamiOracle Complexity Online Learning
Institute for Data, Econometrics, Algorithms, and Learning (IDEAL), hosted by UIC and TTIC · Department of Computer Science, Purdue University
Streaming Adversarial Robustness in Fuzzy ARTMAP: Mechanism-Aligned Evaluation, Progressive Training, and Interpretable Diagnostics
May 7, 2026Shane Cairns, Leonardo Enzo Brito da Silva, Sasha Petrenko +2Adversarial Training Neural Network Robustness
Department of Electrical and Computer Engineering, Missouri University of Science and Technology, Rolla, MO, USA · Instituto Metrópole Digital, Universidade Federal do Rio Grande do Norte, Natal, RN 59078-900, Brazil · Kummer Institute Center for Artificial Intelligence and Autonomous Systems (KICAIAS), Missouri University of Science and Technology, Rolla, MO, USA
MEMOA: Massive Mixtures of Online Agents via Mean-Field Decentralized Nash Equilibria
May 6, 2026Xuwei Yang, David B. Emerson, Fatemeh Tavakoli +1Nash Equilibrium Federated RL
Department of Mathematics and Statistics McMaster University Ontario, Canada · Vector Institute Ontario, Canada
Adaptive Learning Strategies for AoA-Based Outdoor Localization: A Comprehensive Framework
May 6, 2026Bac Trinh-Nguyen, Sara Berri, Sin G. Teo +2Wireless Localization Adaptive Learning
1ETIS UMR 8051, CY Cergy Paris University, ENSEA, CNRS, FR · Institute for Infocomm Research, Agency for Science, Technology and Research (A*STAR), SG · 3Singapore Institute of Technology (SIT), SG +1

Jun 19, 2026J/K move · Enter open · S save

Gradient-Free Warm-Start Library Recovery: an Amortized-Regret Separation

Jianwei Lou

RailMind Systems, Neuss, Germany

Gradient-Free Optimization Continual Learning Online Continual Learning Statistical Learning Theory Online Learning

Open paper AskPDF arXiv

Continual learning that is gradient-free, local, online, and append-only is attractive for edge and streaming deployment, but its value is usually argued informally. We give a provable account on recurring-regime streams. Given segmentation, a warm-start library learner attains amortized recovery cost

O\!\big(KD/\varepsilon^2+(R-K)\logK/Δ^2\big)

versus a memoryless re-estimator's

Θ(RD/\varepsilon^2)

, an advantage

(R-K)\,Θ(D/\varepsilon^2)

growing with dimension

D

and recurrence density. The mechanism is a decoupling: recognizing which of

K

seen regimes is active costs

O(\log K/Δ^2)

, independent of

D

, whereas estimating a regime costs

Θ(D/\varepsilon^2)

. We prove this is tight: matching lower bounds give recognition

Θ(\log K/Δ^2)

and a memoryless-class bound

Ω(RD/\varepsilon^2)

, so each term is individually minimax-tight (the joint statement is conditional). The separation is born-immune (a memoryless learner's advantage is identically zero) and paradigm-level: it matches, and does not beat, a fair spawn-capable Bayesian baseline; the contribution is attaining this cost structure without end-to-end backprop and with zero forgetting by construction. A count-calibrated variant ties the baseline's leading constant up to a bounded, never-negative per-recurrence overshoot, hyperparameter-free and with no per-step transcendentals. We bound the scope: recognizable regimes are capped by simplex packing (walls

e^{Θ(D)}

); autonomous segmentation is impossible at the packing wall (no detector escapes the false-alarm/delay frontier as regimes overlap); the advantage vanishes under overlap. The dimension-dependent separation is corroborated on synthetic streams and real

k

-mer genome distributions (memoryless cost

\propto D^{1.04}

, recognition

D

-independent); the one real sequential stream sits in the

D{=}1

near-null corner.