arXiv research

IDS improves sparse linear bandits by balancing information and regret.

The paper proposes a method to integrate prior information into penalized regression.

New algorithms detect and estimate rank-one signals with prior directional information.

A new algorithm STE for model-based RL improves learning rates.

Bayesian metalearning improves performance in linear bandits with misspecified priors.

PIPA aligns preferences without reinforcement learning, improving language model performance.

Spatial information is not always necessary for spatio-temporal models.

We use the language of uninformative Bayesian prior choice to study the selection of appropriately simple effective models. We advocate for the prior which maximizes the mutual information between parameters and predictions, learning as much as possible from limited data. When many parameters are poorly constrained by …

A fundamental problem in geostatistical modeling is to infer the heterogeneous geological field based on limited measurements and some prior spatial statistics. Semantic inpainting, a technique for image processing using deep generative models, has been recently applied for this purpose, demonstrating its effectiveness…

New bounds on IDS for RL show how to balance computation and learning efficiency.

A significant theoretical advantage of search-and-score methods for learning Bayesian Networks is that they can accept informative prior beliefs for each possible network, thus complementing the data. In this paper, a method is presented for assigning priors based on beliefs on the presence or absence of certain paths …

AutoElicit uses LLMs to quickly create expert priors for predictive models.

Privacy subsidy found in market trading with noisy direction signals.

A novel circuit motif uses sister cells for inference with correlated priors.

Proposes a neural network model to improve predictions in biased datasets.

Proposes a method to identify causal relationships using background knowledge.

Effectively modelling hidden structures in a network is very practical but theoretically challenging. Existing relational models only involve very limited information, namely the binary directional link data, embedded in a network to learn hidden networking structures. There is other rich and meaningful information (e.…

Training of discrete latent variable models remains challenging because passing gradient information through discrete units is difficult. We propose a new class of smoothing transformations based on a mixture of two overlapping distributions, and show that the proposed transformation can be used for training binary lat…

Paper solves NGCA for discrete distributions using LLL method.

Study achieves fairness without demographic info, improving regression tasks.

Unified framework simplifies DPO algorithms for LLM alignment.

Develops a new MCMC-based Wishart prior for Gaussian Process covariance matrix.

Bitcoin option prices reflect both market maker supply and trader demand, especially from those with insider information.

Directed graphs occur throughout statistical modeling of networks, and exchangeability is a natural assumption when the ordering of vertices does not matter. There is a deep structural theory for exchangeable undirected graphs, which extends to the directed case via measurable objects known as digraphons. Using digraph…

Sum-of-Squares lower bound shows NGCA requires more samples than known algorithms.

Bayesian method uses data spectra to estimate non-sparse high-dimensional models.

The inference of deep hierarchical models is problematic due to strong dependencies between the hierarchies. We investigate a specific transformation of the model parameters based on the multivariate distributional transform. This transformation is a special form of the reparametrization trick, flattens the hierarchy a…

Improves AI-prior reliability for Bayesian inference.

Deep learning methods improve subsurface flow modeling efficiency.

New algorithm achieves optimal regret in average reward MDPs without prior bias information.

Improved Monte-Carlo models by constraining mutual information between latent and observable variables.

A new geometric concept, the dead direction, bridges singular learning theory and information geometry.

b-LOAD extends local causal discovery with prior knowledge, improving causal effect estimation.

A new algorithm learns causal relationships from multimodal data.

New method improves Robbins-Monro algorithm convergence with prior information.

One of the most interesting features of Bayesian optimization for direct policy search is that it can leverage priors (e.g., from simulation or from previous tasks) to accelerate learning on a robot. In this paper, we are interested in situations for which several priors exist but we do not know in advance which one fi…

There are three principle paradigms of statistical inference: (i) Bayesian, (ii) information-based and (iii) frequentist inference. We describe an objective prior (the weighting or $w$-prior) which unifies objective Bayes and information-based inference. The $w$-prior is chosen to make the marginal probability an unbia…

New PAC-Bayes method updates priors without losing confidence information.

Additive Bayesian networks are types of graphical models that extend the usual Bayesian generalized linear model to multiple dependent variables through the factorisation of the joint probability distribution of the underlying variables. When fitting an ABN model, the choice of the prior of the parameters is of crucial…

Cold posteriors improve Bayesian neural networks by reducing overestimation of aleatoric uncertainty.

Two data-dependent information metrics are developed to quantify the information of the prior and likelihood functions within a parametric Bayesian model, one of which is closely related to the reference priors from Berger, Bernardo, and Sun, and information measure introduced by Lindley. A combination of theoretical, …

Physics-informed neural networks improve by measuring effective dimensionality of constraints.

Hypermodels improve exploration efficiency and accuracy.

New IDS algorithm refines parameter norm bounds for better bandit performance.

Deep Gaussian Processes model functions on DAGs with partially observed data.

Statsformer validates and adapts LLM-derived semantic priors for improved supervised learning.

The choice of sentence encoder architecture reflects assumptions about how a sentence's meaning is composed from its constituent words. We examine the contribution of these architectures by holding them randomly initialised and fixed, effectively treating them as as hand-crafted language priors, and evaluating the resu…

Develops methods for constructing parameter priors in DAG models.