Adaptive importance sampling is a class of techniques for finding good proposal distributions for importance sampling. Often the proposal distributions are standard probability distributions whose parameters are adapted based on the mismatch between the current proposal and a target distribution. In this work, we present an implicit adaptive importance sampling method that applies to complicated distributions which are not available in closed form. The method iteratively matches the moments of a set of Monte Carlo draws to weighted moments based on importance weights. We apply the method to Bayesian leave-one-out cross-validation and show that it performs better than many existing parametric adaptive importance sampling methods while being computationally inexpensive.

We study the eect of diusing solute atoms on the collective dynamics of dislocations in plastically deforming crystals, by simulating a twodimensional discrete dislocation dynamics model with solute atoms included. We employ various protocols to apply the external stress, including constant, oscillatory and quasistatically increasing stress, and study the resulting dynamics for various values of the solute mobility, temperature, and interaction strength with the dislocations. The values of these parameters dictate if Cottrell clouds are formed around the dislocations, and whether the dislocations are able to drag them along as they move. The relevant solute-induced processes include a temporally increasing average Cottrell cloud size due to cloud merging during the evolution of the dislocation structures subject to constant stresses, and a crossover between a solute-free 'phase' and a regime where solute drag is important for cyclic stresses, controlled by the solute mobility and temperature. Statistics of deformation bursts under quasistatic loading exhibit atypical scaling where the average burst size is directly proportional to its duration, and are also aected by solute-induced strain hardening in the high-stress regime.

Determining the sensitivity of the posterior to perturbations of the prior and likelihood is an important part of the Bayesian workflow. We introduce a practical and computationally efficient sensitivity analysis approach that is applicable to a wide range of models, based on power-scaling perturbations. We suggest a diagnostic based on this that can indicate the presence of prior-data conflict or likelihood noninformativity. The approach can be easily included in Bayesian workflows with minimal work by the model builder. We present the implementation of the approach in our new R package priorsense and demonstrate the workflow on case studies of real data.

No abstract