Develops hedging algorithm for online expert weight allocation with delayed feedback.
problem Adaptive hedging strategies for online expert weight allocation with delayed feedback.
method General Hedging algorithm G \mathcal{G} G based on exponential reweighing of experts' losses. result Proves adversarial loss bounds for the General Hedging algorithm G \mathcal{G} G in the delayed feedback setting. Reinforcement learning improves online matching by combining expert policies.
problem Efficient decision-making in complex systems like cloud services and marketplaces.
method Combines reinforcement learning with expert policies, using advantage-based weight updates.
result The orchestrated policy converges faster and yields higher efficiency than individual experts and conventional RL.
Paper tackles online allocation problems using adversarial training.
problem Online bipartite matching, especially in AdWords.
method Constructs a framework combining game theory, adversarial training, and GANs.
result Designs robust algorithms that perform well under practical and adversarial conditions.
Online L2D algorithm for multiclass classification with varying experts.
problem Handling streaming data, changing expert availability, and shifting expert distribution.
method First online L2D algorithm with O ( ( n + n e ) T 2 / 3 ) O((n+n_e)T^{2/3}) O (( n + n e ) T 2/3 ) and O ( ( n + n e ) T ) O((n+n_e)\sqrt{T}) O (( n + n e ) T ) regret guarantees. result Effective extension of standard L2D to settings with varying expert availability and reliability.
New framework uses tempered optimism to handle imperfect experts in online learning.
problem Challenges of implicit optimism in practical online learning environments.
method Introduces tempered optimism as a framework for online non-convex learning, modifies existing algorithms.
result Demonstrates tempered optimism as a fruitful paradigm for online non-convex learning.
Paper tackles online task allocation in multi-attribute social sensing.
problem Optimized task allocation in dynamic, multi-attribute social sensing.
method Quality-Cost-Aware Online Task Allocation (QCO-TA) scheme using online reinforcement learning.
result Significantly outperforms state-of-the-art baselines in sensing accuracy and cost.
The article improves prediction by aggregating Kalman recursions online.
problem Improving expert aggregation in prediction models.
method Using exponential weights and state-space models to aggregate Kalman recursions.
result New algorithms outperform existing methods in Kalman recursion expert aggregation.
New framework for fair online allocation in continuous time with deadlines.
problem Fair allocation under deadlines in continuous-time online learning.
method Continuous-time utility maximization, dual ascent optimization for time averages.
result Achieves i l d e O ( B − 1 / 2 ) ilde{O}(B^{-1/2}) i l d e O ( B − 1/2 ) regret bound in the absence of statistical knowledge. Adaptive Bayesian learning aggregates experts to improve performance.
problem Bayesian online learning's performance depends on inferential choices.
method Treat Bayesian update rules as experts and aggregate them based on sequential predictive losses.
result The aggregate competes with the best expert in hindsight at a low aggregation cost.
Some online advertising offers pay only when an ad elicits a response. Randomness and uncertainty about response rates make showing those ads a risky investment for online publishers. Like financial investors, publishers can use portfolio allocation over multiple advertising offers to pursue revenue while controlling r…
BOA improves financial forecasting by combining expert models.
problem Challenges in choosing between multiple machine learning models for financial forecasting.
method Online aggregation of expert models using Bernstein Online Aggregation (BOA) procedure.
result BOA leads to better portfolio performance, higher Sharpe Ratio, and lower shortfall.
Optimal online learning for joint pricing and resource allocation.
problem Maximizing net profit in dynamic pricing and resource allocation with stochastic demand.
method Developed an efficient algorithm using a Lower-Confidence Bound (LCB) meta-strategy over multiple OCO agents.
result Achieved i l d e O ( T m n ) ilde{O}(\sqrt{Tmn}) i l d e O ( T mn ) regret, optimal with respect to time horizon T T T . We investigate online classification with paid stochastic experts. Here, before making their prediction, each expert must be paid. The amount that we pay each expert directly influences the accuracy of their prediction through some unknown Lipschitz "productivity" function. In each round, the learner must decide how mu…
One of the major hurdles preventing the full exploitation of information from online communities is the widespread concern regarding the quality and credibility of user-contributed content. Prior works in this domain operate on a static snapshot of the community, making strong assumptions about the structure of the dat…
Paper proposes OPF policy for fair resource allocation with sublinear regret.
problem Fair resource allocation in an online setting against an unrestricted adversary.
method Online Proportional Fair (OPF) policy achieving approximate sublinear regret.
result OPF policy achieves c α c_α c α -approximate sublinear regret with c α ≤ 1.445 c_α \leq 1.445 c α ≤ 1.445 . With the increasing volume of data in the world, the best approach for learning from this data is to exploit an online learning algorithm. Online ensemble methods are online algorithms which take advantage of an ensemble of classifiers to predict labels of data. Prediction with expert advice is a well-studied problem i…
Bayesian framework for online consensus prediction from expert feedback.
problem Online classification with expert consensus prediction, cost-effective.
method General Bayesian framework for dynamic expert consensus estimation.
result Demonstrated superior performance on large-scale crowdsourced datasets.
Novel algorithms for online learning with uncertain feedback graphs reduce regret.
problem Uncertainty in feedback graphs hinders traditional online learning approaches.
method Developed novel online learning algorithms to handle uncertain feedback graphs.
result Proved sublinear regret under mild conditions for the proposed algorithms.
A new mechanism reduces expert belief regret in online forecasting.
problem Minimizing expert belief regret in strategic forecasting.
method Developed a no-regret mechanism for non-myopic experts using online I-ELF.
result Achieved i l d e O ( T N ) ilde{O}(\sqrt{T N}) i l d e O ( T N ) regret for full-information setting. Bayesian algorithms improve online learning with adversaries over infinite action spaces.
problem Online learning with adversaries over infinite action spaces.
method Developed a Thompson sampling algorithm for online learning with an adversary's prior over the space of actions.
result Thompson sampling over a Gaussian process prior achieves a rate of O ( β T d log ( 1 + d λ β ) ) O(β\sqrt{Td\log(1+\sqrt{d}\fracλβ)}) O ( β T d log ( 1 + d β λ ) ) against a β β β -bounded λ λ λ -Lipschitz adversary. Improved time series forecasting with expert loss integration.
problem Enhancing time series forecasting accuracy and efficiency.
method Adaptive Mixture-of-Experts framework with expert-specific loss integration and online learning.
result Significantly improved forecasting accuracy and computational efficiency.
We present an online approach to portfolio selection. The motivation is within the context of algorithmic trading, which demands fast and recursive updates of portfolio allocations, as new data arrives. In particular, we look at two online algorithms: Robust-Exponentially Weighted Least Squares (R-EWRLS) and a regulari…
The paper tackles online learning with expert predictions using only peer feedback.
problem Lack of direct feedback on expert losses in online learning.
method Proposes a peer prediction approach with a carefully designed score function to estimate losses.
result Shows sufficient conditions for bounded regret using the peer score function.
Model explains capital allocation and wealth distribution dynamics in a frictional economy.
problem Understanding capital allocation and wealth distribution dynamics in a frictional economy.
method Mean-field game approach to model interactions between expert and household groups.
result Experts accumulate capital during booms and quickly reverse behavior in busts, even without macro-shocks.
Algorithm allocates perishable resources online to minimize envy and inefficiency.
problem Online allocation of perishable resources to minimize envy and inefficiency.
method Algorithm uses predictions of perishing order and desired envy bound to adaptively allocate resources.
result Algorithm achieves optimal envy-efficiency trade-off as derived from strong lower bounds.
Solves online resource allocation problems with budget constraints.
problem Maximizing revenue for e-commerce platforms under budget constraints.
method Integrated online optimization and learning algorithm for non-stationary Poisson processes.
result Effective and efficient solutions for constrained resource allocation problems.
Existing approaches to resource allocation for nowadays stochastic networks are challenged to meet fast convergence and tolerable delay requirements. The present paper leverages online learning advances to facilitate stochastic resource allocation tasks. By recognizing the central role of Lagrange multipliers, the unde…
New algorithms reduce label collection for online prediction with expert advice.
problem Efficiently predicting binary sequences with expert advice using fewer labels.
method Adaptive selective sampling for exponentially weighted forecasters.
result Label complexity scales roughly as the square root of the number of rounds for a scenario with a strictly better expert.
The paper tackles efficient online learning by achieving minimal regret with respect to the best expert.
problem Achieving minimal regret in online learning problems where the goal is to match the lowest regret of K experts.
method A lazy form of the online subgradient algorithm is used to achieve minimal regret in 'easy' regimes.
result Minimal regret strategies exist for some 'hard' regimes, and the algorithm retains an O ( n ) O(\sqrt{n}) O ( n ) worst-case regret guarantee. The paper tackles online resource allocation with uncertain coefficients and chance constraints.
problem Online stochastic resource allocation problem with chance constraints.
method Linearization and primal-dual algorithms with heuristic corrections.
result Optimality gap and constraint violation are on the order of √n.
Improves online learning with expert demonstrations, quality matters.
problem Improving online learning through offline demonstration data.
method Thompson sampling applied to a multi-armed bandit model, informed by expert demonstrations and Bayes' rule.
result Substantial empirical regret reduction with expert demonstrations, improving online performance.
The paper explores trade-offs between regret and variance in online learning algorithms.
problem Investigating the trade-offs between regret and variance in online learning.
method Analysis of the Exponentially Weighted Average (EWA) algorithm and its variants.
result A variant of EWA either achieves negative regret or guarantees a logarithmic bound on both variance and regret.
New algorithm solves online resource allocation problems efficiently.
problem Dynamic resource allocation in operations research.
method Minimal Selection Principle and MSoE algorithm.
result Ensures optimal cumulative regret bounds in dynamic resource allocation.
New portfolios outperform traditional methods by using factor weights.
problem Improving portfolio allocation in markets driven by factors.
method Factor-weighted Dirichlet portfolios outperform uniform Dirichlet portfolios.
result Factor-weighted portfolios outperform uniformly sampled portfolios in market returns.
Framework for online resource allocation using social welfare functions.
problem Optimal allocation of resources over time steps in a population.
method Confidence sequence framework for SWF-based online learning and inference, valid for any monotonic, concave, and Lipschitz-continuous SWF.
result Achieves near-optimal regret of i l d e O ( n + n k T ) ilde{O}(n+\sqrt{nkT}) i l d e O ( n + nk T ) for SWF-agnostic algorithm SWF-UCB. Improved bounds for online prediction with expert advice.
problem Online prediction with expert advice in finite-horizon games.
method Verification arguments from optimal control theory applied to PDEs to find sub- and supersolutions.
result Explicit bounds for any number of experts and horizon, improving upon previous results.
New algorithm predicts piecewise regular functions online.
problem Online prediction of piecewise regular functions.
method Modified sleeping experts aggregation algorithm.
result Oracle risk bounds for all local regions.
New algorithm optimizes online network resource allocation with long-term constraints.
problem Optimal resource reservation in communication networks with job transfers and budget limits.
method Randomized exponentially weighted method for long-term constraints.
result Upper bound for regret and cumulative constraint violations established.
Boosting improves online decision-making for large expert sets.
problem Online convex optimization with many experts is infeasible.
method Generalizes online boosting to online convex optimization and bandit linear optimization settings.
result Near-optimal regret guarantees for various feedback models.
A fast method combines deep mixtures of sparse GPs for flexible modeling.
problem Flexible modeling with changing output densities.
method Designing gating network with DNN for selecting sparse GPs, using CCR algorithm.
result The method outperforms competing methods in accuracy and uncertainty quantification.
Algorithm detects changes online using expert tracking.
problem Online change point detection in nonparametric settings.
method Sequential score function estimation and tracking the best expert approach.
result Algorithm performs well in artificial and real-world data.
Private learning can be used to efficiently solve online learning problems.
problem The relationship between differentially private learning and online learning efficiency.
method Derive an efficient black-box reduction from differentially private learning to online learning from expert advice.
result An efficient differentially private learner implies an efficient online learner.
This work explains why online imitation learning improves faster than theory predicts.
problem Online imitation learning's empirical policy improvement speed exceeds theoretical predictions.
method The authors analyze online imitation learning with a convex, smooth, and non-negative loss function, proving policy improvement in expectation and high probability.
result Adopting a sufficiently expressive policy class in online IL increases both policy improvement speed and performance bias.
Two adaptive algorithms improve tracking regret in dynamic expert advice problems.
problem Prediction with expert advice in dynamic environments.
method Developed two adaptive and efficient algorithms using online mirror descent framework.
result Achieved data-dependent tracking regret bounds for both algorithms.
Optimizes crowdsourced preference-based subjective evaluation with online learning.
problem Large-scale evaluation of generative media using crowdsourcing due to combinatorial explosion.
method Automatic optimization of pair combination selections and evaluation volumes with online learning.
result Optimizes evaluation by reducing pair combinations and allocating optimal evaluation volumes.
New algorithm reduces expert prediction regret for two experts.
problem Efficient prediction with two experts under fixed time constraints.
method Optimal algorithm based on stochastic calculus techniques.
result Achieves optimal regret of sqrt(T/2π) + O(1) with O(1) per-turn processing time.
The paper introduces SuccessProbaMax to optimize policy success probability in online advertising.
problem Optimizing policy success probability in online advertising systems.
method SuccessProbaMax algorithm that optimizes for the probability of success rather than expected value.
result SuccessProbaMax outperforms conventional algorithms in terms of success rate.
The paper tackles budget allocation for multiple campaigns using a novel combinatorial bandit approach.
problem Maximizing cumulative returns with limited budgets across various ad lines.
method Formulated as a multi-task combinatorial bandit problem, integrates Bayesian hierarchical models, and uses Thompson sampling.
result Demonstrates robustness and adaptability in maximizing overall cumulative returns.