CTS reduces regret in probabilistically triggered combinatorial bandits.
problem Optimizing decisions with probabilistically triggered arms in combinatorial multi-armed bandits.
method Combinatorial Thompson Sampling (CTS) with a regret bound analysis.
result Derives an O ( ∑ i = 1 m log T / ( p i Δ i ) ) O(\sum_{i =1}^m \log T / (p_i Δ_i)) O ( ∑ i = 1 m log T / ( p i Δ i )) regret bound for CTS. New algorithm for contextual combinatorial bandits with probabilistic arm triggering.
problem Optimizing decisions in dynamic environments with probabilistic arm availability.
method C^2-UCB-T and VAC^2-UCB algorithms with TPM and VM conditions.
result Achieved improved regret bounds for contextual combinatorial bandits.
Improved UCB and Thompson Sampling policies for CMAB with probabilistically triggered arms achieve bounded regret.
problem Combinatorial multi-armed bandit problem with probabilistically triggered arms.
method Upper Confidence Bound (UCB) policies and Combinatorial Thompson Sampling (CTS).
result CUCB- κ κ κ and CTS achieve O ( T ) O(\sqrt{T}) O ( T ) gap-independent regret. Paper improves CMAB regret bounds by reducing batch-size dependency.
problem Reducing batch-size dependency in combinatorial semi-bandits.
method Developed BCUCB-T and SESCB algorithms with new TPVM conditions.
result Significantly improved regret bounds for various applications.
We study combinatorial multi-armed bandit with probabilistically triggered arms (CMAB-T) and semi-bandit feedback. We resolve a serious issue in the prior CMAB-T studies where the regret bounds contain a possibly exponentially large factor of 1 / p ∗ 1/p^* 1/ p ∗ , where p ∗ p^* p ∗ is the minimum positive probability that an arm is trigg…
CTS improves combinatorial optimization in unknown environments.
problem Optimizing actions from a large set of alternatives in unknown environments.
method Combinatorial Thompson Sampling (CTS) for semi-bandit feedback.
result CTS achieves optimal regret bounds in various networking problems.
Graph-Triggered Bandits unify rested and restless bandits with graph-defined arm interactions.
problem Modeling sequential decision-making problems with evolving arm rewards.
method Graph-Triggered Bandits (GTBs) framework that generalizes rested and restless bandits using a graph.
result Rested and restless bandits are special cases of GTBs for suitable graphs.
Study improves CTS's approximation regret for combinatorial bandits.
problem Improving CTS's performance on non-exact oracles.
method Develops a new O ( log ( T ) / Δ ) \mathcal{O}(\log(T)/Δ) O ( log ( T ) /Δ ) upper bound for CTS under specific conditions. result First O ( log ( T ) / Δ ) \mathcal{O}(\log(T)/Δ) O ( log ( T ) /Δ ) approximation regret upper bound for CTS. The paper analyzes online learning with probabilistic graph feedback, matching regret bounds with high probability.
problem Online learning with probabilistic graph feedback, covering both one-step and cascade cases.
method Analyzed asymptotic lower bounds and designed algorithms for both cases, matching regret bounds with high probability.
result Regret upper bounds match asymptotic lower bounds with high probability.
ET-GP-UCB optimizes time-varying functions without knowing change rates.
problem Sequentially optimizing a time-varying objective function with unknown change rates.
method Event-triggered Bayesian optimization with adaptive resets based on probabilistic uniform error bounds.
result ET-GP-UCB outperforms other GP-UCB algorithms in synthetic and real-world data.
New algorithm minimizes regret in multi-agent bandit problem with probabilistic communication.
problem Minimizing group regret in multi-agent multi-armed bandit problem with probabilistic communication.
method Proposes a new UCB-based algorithm for decentralized multi-agent multi-armed bandit problem on d d d -regular graphs with probabilistic communication. result The proposed algorithm outperforms state-of-the-art algorithms in minimizing group regret.
New algorithm for multi-player bandits with selfish players, achieving logarithmic regret.
problem Challenges of robustness to selfish players in multi-player bandits.
method First algorithm robust to selfish players achieving logarithmic regret, with or without collision observation.
result Achieved logarithmic regret for robust algorithms to selfish players in multi-player bandits.
New algorithm tackles non-stationary combinatorial semi-bandit problems with optimal regret bounds.
problem Non-stationary combinatorial semi-bandit problems in switching and dynamic environments.
method Developed algorithms for both switching and dynamic cases, achieving nearly optimal regret bounds.
result Achieved nearly optimal regret bounds in both switching and dynamic cases.
Combines offline and online learning for identifying the best arm in bandits.
problem Identifying the best arm in stochastic bandits with offline data.
method Lower bound analysis and algorithm development for optimal best-arm identification.
result Developed algorithms matching lower bound on sample complexity for small δ.
We present a probabilistic model of events in continuous time in which each event triggers a Poisson process of successor events. The ensemble of observed events is thereby modeled as a superposition of Poisson processes. Efficient inference is feasible under this model with an EM algorithm. Moreover, the EM algorithm …
New Thompson Sampling for partially observed context bandits reduces regret logarithmically with time.
problem Improving Thompson Sampling for partially observed context bandits.
method Proposed a Thompson Sampling algorithm for partially observable contextual multi-armed bandits with theoretical performance guarantees.
result Regret scales logarithmically with time and the number of arms, and linearly with the dimension.
Motivated by models of human decision making proposed to explain commonly observed deviations from conventional expected value preferences, we formulate two stochastic multi-armed bandit problems with distorted probabilities on the reward distributions: the classic K K K -armed bandit and the linearly parameterized bandit…
MINTS uses a minimalist Bayesian framework to tackle multi-armed bandits with structural constraints.
problem Sequential decision-making under uncertainty with complex structural constraints.
method Minimalist Bayesian framework with profile likelihood to eliminate nuisance parameters.
result MINTS achieves near-optimal regret guarantees and adapts to unimodal structure.
A new Bayesian framework simplifies stochastic optimization by focusing on key parameters.
problem Bayesian methods struggle with complex structural constraints.
method Minimalist Bayesian framework that eliminates nuisance parameters via profile likelihood.
result Near-optimal regret guarantees for multi-armed bandits and convex optimization.
Study best arm identification with contextual info, achieving optimal misidentification probability.
problem Identify the best treatment arm with minimal misidentification probability in a small gap scenario.
method Developed RS-AIPW strategy that matches lower bound of misidentification probability in the small-gap regime.
result RS-AIPW strategy is asymptotically optimal for best arm identification.
Study on reward poisoning attacks on CMAB, revealing attackability depends on adversary's knowledge.
problem Reward poisoning attacks on Combinatorial Multi-Armed Bandits (CMAB).
method Provided a sufficient and necessary condition for attackability, devised an attack algorithm.
result Attackability of CMAB depends on adversary's knowledge of the bandit instance.
Paper improves voice trigger detection for privacy-centric smart assistants.
problem Mitigating false triggers in voice-activated smart assistants.
method Analyzing ASR lattices using graph neural networks (GNN).
result GNNs effectively reduce false triggers by ~87% at 99% true positive rate.
Pricing Chinese convertible bonds using Monte Carlo simulation and dynamic programming.
problem Pricing Chinese convertible bonds accurately.
method Monte Carlo simulation and dynamic programming with regression and backward induction.
result An underpriced strategy significantly outperforms benchmarks.
Event-triggered learning reduces communication in networked control systems.
problem Reduction of communication in networked control systems.
method Triggered learning experiments when communication performance is poor, using statistical properties of inter-communication times.
result Event-triggered learning improves robustness and communication efficiency.
New method defends against neural backdoors using generative modeling.
problem Neural backdoor attacks pose a significant security threat to deep learning models.
method Proposes max-entropy staircase approximator (MESA) for high-dimensional sampling-free generative modeling of backdoor trigger distributions.
result Demonstrates the effectiveness of MESA in modeling backdoor trigger distributions and robustness of the proposed defense method.
New probabilistic deep learning models using random SPNs are robust and interpretable.
problem Inference limitations in probabilistic deep learning models.
method Random Sum-Product Networks (RAT-SPNs) trained with deep learning techniques.
result RAT-SPNs yield comparable predictions to deep neural networks with interpretability and robustness.
Paper solves time-inconsistent control problems with BSDEs.
problem Time-inconsistent stochastic control in continuous time.
method Probabilistic representation via BSDEs.
result Equilibrium value function resolved for inconsistent cases.
MISA detects Trojan triggers in neural networks at inference time.
problem Trojan attacks on neural networks that respond to specific trigger patterns.
method MISA uses misattributions to detect anomalous feature activations.
result MISA achieves 96% AUC in detecting Trojan triggers without assumptions.
New graph feedback model for bandits with improved regret bounds.
problem Understanding how graph structure affects regret in bandit problems.
method Introduced fractional weak domination number and k k k -packing independence number to capture upper and lower bounds on regret. Used strong duality theorem to derive upper and lower bounds. result Proved general upper and lower bounds on regret for various graph structures, showing tightness up to a logarithmic factor.
Automatically improves Monte Carlo estimators in probabilistic programs.
problem Reducing variance in Monte Carlo estimators for probabilistic programs.
method Dynamic mechanism using conjugate priors and affine transformations.
result Automatic Rao-Blackwellization and locally-optimal proposals.
Unified framework for distributional regret in bandits and reinforcement learning.
problem Characterizing the distribution of regret in multi-armed bandits and reinforcement learning.
method Unified framework with a UCBVI-style algorithm and distributional regret bounds.
result Distributional regret bounds with optimal trade-offs between expected and distributional regret.
PHAZE framework uses zkML and hashing for fast, verifiable LHC trigger decisions.
problem Inefficient inference on large machine learning models for LHC trigger performance.
method Cryptographic techniques like hashing and zkML for low latency, certifiable inference.
result Achieves nanosecond-order latency for LHC triggers, enabling dynamic low-level triggers.
Improved voice trigger detection in noisy environments.
problem Complex acoustic environments and lack of trigger phrase training data.
method Two-stage cascaded architecture with multi-task learning.
result Model reduces errors by half compared to baseline in challenging conditions.
S-TRIGGER learns state representations for continual learning.
problem Efficiently compress and maintain past knowledge in changing environments.
method Generative Replay with self-triggered environment change detection.
result S-TRIGGER enables fast and high-performing Reinforcement Learning without catastrophic forgetting.
PPPD framework extracts physical characterizations from stochastic mechanical systems.
problem Complex system behavior requires more than probabilistic descriptions of QoI.
method Probabilistic Performance-Pattern Decomposition (PPPD) framework.
result Decomposes system behaviors into meaningful patterns in response space.
CascadeBAI identifies best arms in cascading bandits with fixed confidence.
problem Finding the best set of items in cascading bandits with limited feedback.
method Developed CascadeBAI algorithm, derived upper and lower bounds on time complexity, introduced left-sided sub-Gaussian random variables.
result CascadeBAI is optimal in some practical regimes and performs well with limited feedback.
Improved batch-size independent regret bounds for nonlinear reward functions.
problem Nonlinear reward functions in combinatorial multi-armed bandit problems.
method Introducing Gini-weighted smoothness to account for both nonlinearity and concentration properties of arms.
result Achieved dramatic improvements in upper bounds for the probabilistic maximum coverage problem.
Thompson Sampling improves decision-making in partially observed contexts.
problem Balancing exploration and exploitation in partially observed contextual bandits.
method Thompson Sampling policy for learning optimal arms from noisy linear functions of unobserved context vectors.
result Thompson Sampling achieves poly-logarithmic regret and square-root consistency of parameter estimation.
New strategy for DOE tasks using posterior sampling and probabilistic programming.
problem Sequential data collection for specific goals.
method Myopic Posterior Sampling (MPS) inspired by Thompson sampling.
result Competitive with specialised methods and applicable to complex tasks.
LinConTS improves regret and constraint violations in probabilistic linearly constrained bandits.
problem Maximizing cumulative reward under probabilistic linear constraints.
method LinConTS, a Thompson Sampling-based algorithm for bandits with linear constraints.
result LinConTS achieves O(log T) regret and constraint violations for suboptimal arms.
We deliver a call to arms for probabilistic numerical methods: algorithms for numerical tasks, including linear algebra, integration, optimization and solving differential equations, that return uncertainties in their calculations. Such uncertainties, arising from the loss of precision induced by numerical calculation …
Backdoor attacks make models predict a specific class near triggers, smoothing their decision function.
problem Understanding and mitigating backdoor attacks on deep neural networks.
method Defined a measure to quantify backdoor smoothing and detected other smoothing patterns.
result Backdoor attacks induce a smoother decision function around triggered samples.
The paper studies how and when a treatment triggers different effects for individuals.
problem Estimating how treatment effects vary among individuals based on their characteristics.
method Tree-based learning method to find individual-level treatment triggers.
result The proposed method learns treatment triggers better than existing approaches.
SPARQ-SGD optimizes communication in decentralized SGD with event-triggered and compressed updates.
problem Efficient communication in decentralized stochastic optimization for large-scale models.
method Event-triggered and compressed algorithm with quantized and sparsified model parameters.
result SPARQ-SGD converges with efficiency comparable to uncompressed training, demonstrating significant communication savings.
TS-Insight visualizes Thompson Sampling for better debugging and trust.
problem Thompson Sampling's black box nature hinders debugging and trust.
method TS-Insight is a visual analytics tool that traces evolving posteriors and evidence counts.
result Visualizations help in verifying, diagnosing, and explaining Thompson Sampling dynamics.
BadGD identifies gradient descent vulnerabilities through strategic backdoor attacks.
problem Gradient descent vulnerabilities through malicious data manipulation.
method Introduces Max RiskWarp, Max GradWarp, and Max GradDistWarp triggers to exploit gradient descent.
result Demonstrates how malicious triggers can significantly alter loss landscapes and gradient calculations.
Corporate defaults may be triggered by some major market news or events such as financial crises or collapses of major banks or financial institutions. With a view to develop a more realistic model for credit risk analysis, we introduce a new type of reduced-form intensity-based model that can incorporate the impacts o…
Evolutionary algorithm improves DNN watermarking with fewer false positives.
problem Protecting deep learning models from piracy and proving ownership.
method Evolutionary algorithm for generating and optimizing trigger patterns.
result Reduces false positive rates in DNN watermarking.