A new framework enables real-time task trade-off control.
problem Conflict between multiple related tasks in a fixed model capacity.
method Formulates MTL as a preference-conditioned multiobjective optimization problem; uses a hypernetwork-based neural network.
result A single model can handle different trade-off preferences among multiple tasks.
Bayesian optimization offers the possibility of optimizing black-box operations not accessible through traditional techniques. The success of Bayesian optimization methods such as Expected Improvement (EI) are significantly affected by the degree of trade-off between exploration and exploitation. Too much exploration c…
MPC outperforms reactive budgeting in non-stationary return environments.
problem Optimizing budget allocation under non-stationary returns.
method Receding-horizon Model Predictive Control (MPC) compared to reactive policies.
result MPC consistently outperforms reactive budgeting when return dynamics are predictable.
DPA aligns LLMs with multi-objective rewards for diverse user preferences.
problem Fine-grained control over LLMs for diverse user needs.
method Integrates multi-objective reward modeling and directional preference control.
result DPA offers better performance trade-offs and intuitive user control over LLM generation.
New insights into RL efficiency from managing time discretization.
problem The impact of time discretization on RL methods in continuous-time systems.
method Analysis of Monte-Carlo policy evaluation for LQR systems.
result An optimal choice of temporal resolution for a given data budget improves policy evaluation efficiency.
Dealing with high variance is a significant challenge in model-free reinforcement learning (RL). Existing methods are unreliable, exhibiting high variance in performance from run to run using different initializations/seeds. Focusing on problems arising in continuous control, we propose a functional regularization appr…
We introduce Dynamic Deep Neural Networks (D2NN), a new type of feed-forward deep neural network that allows selective execution. Given an input, only a subset of D2NN neurons are executed, and the particular subset is determined by the D2NN itself. By pruning unnecessary computation depending on input, D2NNs provide a…
Gradient-based methods for optimisation of objectives in stochastic settings with unknown or intractable dynamics require estimators of derivatives. We derive an objective that, under automatic differentiation, produces low-variance unbiased estimators of derivatives at any order. Our objective is compatible with arbit…
WHOMP optimizes randomized controlled trials by minimizing subgroup bias.
problem Minimizing subgroup bias in randomized controlled trials.
method Wasserstein Homogeneity Partition (WHOMP) method.
result WHOMP optimally minimizes type I and type II errors in trials.
Learning data representations that are transferable and are fair with respect to certain protected attributes is crucial to reducing unfair decisions while preserving the utility of the data. We propose an information-theoretically motivated objective for learning maximally expressive representations subject to fairnes…
We address the problem of estimating the inputs of a dynamical system from measurements of the system's outputs. To this end, we introduce a novel estimation algorithm that explicitly trades off bias and variance to optimally reduce the overall estimation error. This optimal trade-off is done efficiently and adaptively…
New RL algorithms improve control tasks with data reuse.
problem Real-world control requires performance guarantees and data efficiency.
method Generalized Policy Improvement combining on-policy guarantees and sample reuse.
result Extensive experimental analysis shows benefits of new algorithms.
We study Bayesian optimal control of a general class of smoothly parameterized Markov decision problems. Since computing the optimal control is computationally expensive, we design an algorithm that trades off performance for computational efficiency. The algorithm is a lazy posterior sampling method that maintains a d…
A great variety of off-policy learning algorithms exist in the literature, and new breakthroughs in this area continue to be made, improving theoretical understanding and yielding state-of-the-art reinforcement learning algorithms. In this paper, we take a unifying view of this space of algorithms, and consider their t…
This paper analyzes bias-variance trade-off for clipped SFOMs, improving complexity guarantees for heavy-tailed noise.
problem Improving complexity guarantees for stochastic optimization methods with heavy-tailed noise.
method Novel analysis of bias-variance trade-off in gradient clipping for clipped SFOMs.
result Improved complexity guarantees for clipped SFOMs across various tail indices, including infinite mean noise.
Control of non-episodic, finite-horizon dynamical systems with uncertain dynamics poses a tough and elementary case of the exploration-exploitation trade-off. Bayesian reinforcement learning, reasoning about the effect of actions and future observations, offers a principled solution, but is intractable. We review, then…
One important partition of algorithms for controlling the false discovery rate (FDR) in multiple testing is into offline and online algorithms. The first generally achieve significantly higher power of discovery, while the latter allow making decisions sequentially as well as adaptively formulating hypotheses based on …
EDU method finds diverse optimal solutions for expensive simulators.
problem Optimizing expensive black-box simulators for diverse solutions.
method EDU method searches for diverse locally-optimal solutions within a tolerance level.
result EDU yields a closed-form acquisition function facilitating efficient sequential queries.
We consider the exploration-exploitation tradeoff in linear quadratic (LQ) control problems, where the state dynamics is linear and the cost function is quadratic in states and controls. We analyze the regret of Thompson sampling (TS) (a.k.a. posterior-sampling for reinforcement learning) in the frequentist setting, i.…
Paper develops a dynamic Bayesian approach for active learning that optimizes exploration-exploitation balance.
problem Balancing exploration and exploitation in active learning for unknown functions.
method Develops BHEEM, a Bayesian hierarchical approach with approximate Bayesian computation for sampling trade-off parameters.
result BHEEM achieves at least 21% and 11% improvement over pure exploration and exploitation strategies respectively.
Iterative method learns unknown constraints for MPC control.
problem Learning to satisfy unknown polyhedral state constraints in iterative MPC.
method Collects and improves estimates of unknown constraints using collected data, designs an MPC controller to satisfy the estimated constraints.
result Robust and probabilistic guarantees of constraint satisfaction as a function of task iterations.
Improved control approach for correlated bandits with better performance.
problem General multi-armed bandit problem with correlated elements.
method Introducing entropy regularisation to obtain a smooth asymptotic approximation of the value function, leading to a semi-index approximation of the optimal decision process.
result Performance of Asymptotic Randomised Control (ARC) algorithm compares favorably with other approaches.
GCVAE improves disentanglement in VAEs while balancing reconstruction error.
problem Improving disentanglement in VAEs while maintaining low reconstruction error.
method Introduces three controllable Lagrangian hyperparameters to optimize reconstruction and KL divergence loss.
result GCVAE outperforms state-of-the-art models in disentanglement while balancing reconstruction.
EHVI outperforms scalarized EI in MOBO for molecule design.
problem Benchmarking MOBO strategies for molecule design.
method Compared EHVI against fixed-weight scalarized EI in MOBO.
result EHVI consistently outperforms scalarized EI in molecular optimization tasks.
Bayesian imputation optimizes bias-variance tradeoff in time-series data.
problem Look-ahead bias in imputation of missing time-series data.
method Wasserstein interpolation for Bayesian posterior consensus distribution.
result Optimal control of look-ahead bias and variance in imputation.
The paper proposes a neural network method to estimate treatment effects by balancing treated and control distributions.
problem Estimating individual and average treatment effects from observational data.
method Balance regularization of multi-head neural network architectures to reduce confounding effects.
result The approach reduces bias-variance trade-off and improves treatment effect estimation.
Study nonparametric estimator for Markov chain transition matrices in offline setting.
problem Estimating transition matrices of finite controlled Markov chains from logged data.
method Developed sample complexity bounds and conditions for minimaxity.
result Achieving certain statistical risk requires balancing mixing properties and sample size.
Model analyzes OTC market making with reputation feedback.
problem Optimizing electronic OTC liquidity provision considering reputation.
method Developed a stochastic-control model with feedback loops.
result Policy alternates between reputation-building and franchise monetization phases.
Model analyzes how reputation feedback affects OTC market making.
problem Understanding and optimizing OTC market making strategies.
method Developed a stochastic-control model with feedback loops.
result Policy alternates between reputation building and franchise monetization.
Framework audits synthetic datasets for trustworthiness across various use cases.
problem Assessing the trustworthiness of synthetic datasets and models.
method Holistic auditing framework focusing on bias, fidelity, utility, robustness, and privacy.
result Introduces a trustworthiness index and model selection process for controllable trade-offs.
Study tests whether trade-off functions are above or below benchmarks using finite samples.
problem Testing trade-off functions between unknown distributions.
method Identifies a condition for nontrivial testing, constructs a test with error guarantees, and inverts the test for confidence bands.
result Finite-sample testing is possible under specific structural assumptions about rejection regions.
VAE struggles with distribution class sharpness, which can be learned dynamically.
problem VAE struggles with distribution class sharpness, leading to blurry reconstructions.
method Established that VAE fails if distribution class sharpness does not match data scale. Suggested learning distribution sharpness dynamically.
result Dynamic adjustment of distribution sharpness improves VAE performance, escaping local optima.
We consider an agent who is involved in a Markov decision process and receives a vector of outcomes every round. Her objective is to maximize a global concave reward function on the average vectorial outcome. The problem models applications such as multi-objective optimization, maximum entropy exploration, and constrai…
This paper analyzes error in SKI for Gaussian Processes, providing conditions for linear time inference.
problem Lack of rigorous theoretical error analysis for SKI.
method Proved error bounds for SKI Gram matrix, examined error effects, provided practical guidelines.
result Identified two dimensionality regimes for SKI's scalability-accuracy trade-offs.
New diffusion models improve counterfactual image generation with semantic control.
problem Challenges in preserving identity, maintaining quality, and ensuring causal model faithfulness in counterfactual image generation.
method Integrates semantic representations into diffusion models through Pearlian causality, introducing spatial, semantic, and dynamic abduction.
result Demonstrates high-level semantic identity preservation and principled trade-offs between faithful causal control and identity preservation.
In many real-world applications of Machine Learning it is of paramount importance not only to provide accurate predictions, but also to ensure certain levels of robustness. Adversarial Training is a training procedure aiming at providing models that are robust to worst-case perturbations around predefined points. Unfor…
The Information Bottleneck (IB) objective uses information theory to formulate a task-performance versus robustness trade-off. It has been successfully applied in the standard discriminative classification setting. We pose the question whether the IB can also be used to train generative likelihood models such as normal…
DynamicVAE improves disentanglement and reconstruction accuracy without sacrificing one for the other.
problem The inherent trade-off between disentanglement and reconstruction accuracy in VAE models.
method DynamicVAE uses a modified incremental PI controller to dynamically adjust the weight β during training, decoupling disentanglement and reconstruction accuracy.
result DynamicVAE significantly improves reconstruction accuracy while maintaining disentanglement comparable to existing methods.
Finite resources limit false discovery rate control in structured hypothesis spaces.
problem Controlling false discovery rate in hypothesis testing with finite data and structured hypothesis spaces.
method Framework for exact FDR control and adaptive power maximization.
result Exact FDR control and adaptive power maximization.
The trade-off between the cost of acquiring and processing data, and uncertainty due to a lack of data is fundamental in machine learning. A basic instance of this trade-off is the problem of deciding when to make noisy and costly observations of a discrete-time Gaussian random walk, so as to minimise the posterior var…
Optimizes machine learning models while controlling risks.
problem Finding a model configuration that balances multiple conflicting metrics.
method Combines Bayesian Optimization with rigorous risk-controlling procedures.
result Identifies and selects Pareto optimal configurations with guaranteed risk levels.
Gradient descent performs well on weakly convex losses, offering generalization guarantees.
problem Learning with weakly convex losses using gradient descent.
method Analyzing the stability of gradient descent through the smallest eigenvalue of the Hessian.
result Generalization error bounds hold under a wider range of step sizes.
SVM used for estimating treatment effects without confounding.
problem Estimating average treatment effects in the presence of confounding variables.
method Adapts SVM classifier as a kernel-based weighting procedure to balance covariates and estimate causal effects.
result SVM provides a continuous relaxation of the quadratic integer program for balancing covariates and maximizing effective sample size.
CausalMix generates synthetic data with causal controls for mixed-type tables.
problem Synthetic data for causal inference with mixed-type and multimodal tabular data.
method CausalMix combines Gaussian latent priors with data-type-specific decoders for control over overlap, confounding, and treatment effect heterogeneity.
result CausalMix achieves state-of-the-art distributional metrics and stable causal control.
New method controls posterior collapse in VAEs without network architecture constraints.
problem Posterior collapse in VAEs reduces diversity of generated samples.
method Introduces Latent Reconstruction (LR) loss to control posterior collapse.
result Controls posterior collapse on various datasets without architectural constraints.
Model selection on validation data is an essential step in machine learning. While the mixing of data between training and validation is considered taboo, practitioners often violate it to increase performance. Here, we offer a simple, practical method for using the validation set for training, which allows for a conti…
The paper analyzes the benefit-cost ratio for feature selection in machine learning.
problem Tackling the challenge of distinguishing relevant features from noise in feature selection.
method Simulation study with different cost and data settings to analyze the benefit-cost ratio.
result The benefit-cost ratio can overemphasize cheap noise features in scenarios with large cost differences and small effect sizes.
New theory explains how overparametrized neural networks generalize well without bias-variance trade-off.
problem Overparametrized neural networks generalize well despite classical bias-variance trade-off.
method Nonasymptotic generalization theory for two-layer neural networks with ReLU activation, incorporating scaled variation regularization.
result Prediction bounds for all network widths reproduce the double descent phenomenon, and overparametrized models are nearly minimax optimal.