arXiv subjects

Erik Štrumbelj

Publications and source records attributed to Erik Štrumbelj.

10 recordsLinked to original sources

arXiv2026

A Scoping Review of the Negative Effects of Digital Technology on Cognition

The rapid integration of digital technology into daily life has prompted sustained concern regarding its impact on human cognition. To characterize documented negative effects and the conditions under which they arise, we conducted a scoping review isolating the documented negative effects of digital technology use on cognition. Using a hybrid automated and manual search strategy, we identified foundational seed papers via Scopus and executed an algorithmic citation snowballing process via the OpenAlex API to capture relevant empirical and non-empirical literature. The resulting synthesis of 937 papers (584 empirical, 353 non-empirical) spans legacy screens, multitasking, smartphones, and the nascent work on generative artificial intelligence (AI). Evidence suggests an evolution in the nature of cognitive risk: while research on earlier technologies predominantly describes disruptions to resource allocation, early findings on AI point toward a hypothesized erosion of higher-order cognition. We analyze these risks across cognitive domains through four mechanisms: functional interference, neurochemical dysregulation, structural neuroplasticity, and psychosocial displacement. Effects are frequently moderated by socioeconomic status and environmental factors, suggesting that cognitive decline is often mediated by the displacement of activities rather than direct technological toxicity. Finally, the paper examines how habitual digital offloading could theoretically deplete cognitive reserve, creating downstream risks for long-term health. The collective evidence suggests an efficiency-atrophy paradox, where digital tools optimize short-term performance at the potential expense of long-term unassisted cognition.

cs.CY↗

arXiv2026

A General Approach to Visualizing Uncertainty in Statistical Graphics

We present a general approach to visualizing uncertainty in static 2-D statistical graphics. If we treat a visualization as a function of its underlying quantities, uncertainty in those quantities induces a distribution over images. We show how to aggregate these images into a single visualization that represents the uncertainty. The approach can be viewed as a generalization of sample-based approaches that use overlay. Notably, standard representations, such as confidence intervals and bands, emerge with their usual coverage guarantees without being explicitly quantified or visualized. As a proof of concept, we implement our approach in the IID setting using resampling, provided as an open-source Python library. Because the approach operates directly on images, the user needs only to supply the data and the code for visualizing the quantities of interest without uncertainty. Through several examples, we show how both familiar and novel forms of uncertainty visualization can be created. The implementation is not only a practical validation of the underlying theory but also an immediately usable tool that can complement existing uncertainty-visualization libraries.

stat.ME↗

arXiv2025

Reducing normalizing flow complexity for MCMC preconditioning

Preconditioning is a key component of MCMC algorithms that improves sampling efficiency by facilitating exploration of geometrically complex target distributions through an invertible map. While linear preconditioners are often sufficient for moderately complex target distributions, recent work has explored nonlinear preconditioning with invertible neural networks as components of normalizing flows (NFs). However, empirical and theoretical studies show that overparameterized NF preconditioners can degrade sampling efficiency and fit quality. Moreover, existing NF-based approaches do not adapt their architectures to the target distribution. Related work outside of MCMC similarly finds that suitably parameterized NFs can achieve comparable or superior performance with substantially less training time or data. We propose a factorized preconditioning architecture that reduces NF complexity by combining a linear component with a conditional NF, improving adaptability to target geometry. The linear preconditioner is applied to dimensions that are approximately Gaussian, as estimated from warmup samples, while the conditional NF models more complex dimensions. Our method yields significantly better tail samples on two complex synthetic distributions and consistently better performance on a sparse logistic regression posterior across varying likelihood and prior strengths. It also achieves higher effective sample sizes on hierarchical Bayesian model posteriors with weak likelihoods and strong funnel geometries. This approach is particularly relevant for hierarchical Bayesian model analyses with limited data and could inform current theoretical and software strides in neural MCMC design.

cs.LG↗

arXiv2025

Quantifying Uncertainty: All We Need is the Bootstrap?

A critical literature review and comprehensive simulation study is used to show that (a) non-parametric bootstrap is a viable alternative to commonly taught and used methods in basic estimation tasks (mean, variance, quartiles, correlation) and (b), contrary to recommendations in most related work, double bootstrap performs better than BCa. Quantifying uncertainty through standard errors, confidence intervals, hypothesis tests, and related measures is a fundamental aspect of statistical practice. However, these techniques involve a variety of methods, mathematical formulas, and underlying concepts, which can be complex. Could the non-parametric bootstrap, known for its simplicity and general applicability, serve as a universal alternative? This paper addresses this question through a review of the existing literature and a simulation analysis of one- and two-sided confidence intervals across varying sample sizes, confidence levels, data-generating processes, and statistical functionals. Results show that the double bootstrap consistently performs best and is a promising alternative to traditional methods used for common statistical tasks. These results suggest that the bootstrap, particularly the double bootstrap, could simplify statistical education and practice without compromising effectiveness.

stat.ME↗

arXiv2025

Empirical evaluation of normalizing flows in Markov Chain Monte Carlo

Recent advances in MCMC use normalizing flows to precondition target distributions and enable jumps to distant regions. However, there is currently no systematic comparison of different normalizing flow architectures for MCMC. As such, many works choose simple flow architectures that are readily available and do not consider other models. Guidelines for choosing an appropriate architecture would reduce analysis time for practitioners and motivate researchers to take the recommended models as foundations to be improved. We provide the first such guideline by extensively evaluating many normalizing flow architectures on various flow-based MCMC methods and target distributions. When the target density gradient is available, we show that flow-based MCMC outperforms classic MCMC for suitable NF architecture choices with minor hyperparameter tuning. When the gradient is unavailable, flow-based MCMC wins with off-the-shelf architectures. We find contractive residual flows to be the best general-purpose models with relatively low sensitivity to hyperparameter choice. We also provide various insights into normalizing flow behavior within MCMC when varying their hyperparameters, properties of target distributions, and the overall computational budget.

cs.LG↗

arXiv2025

RelDiff: Relational Data Generative Modeling with Graph-Based Diffusion Models

Real-world databases are predominantly relational, comprising multiple interlinked tables that contain complex structural and statistical dependencies. Learning generative models on relational data has shown great promise in generating synthetic data and imputing missing values. However, existing methods often struggle to capture this complexity, typically reducing relational data to conditionally generated flat tables and imposing limiting structural assumptions. To address these limitations, we introduce RelDiff, a novel diffusion generative model that synthesizes complete relational databases by explicitly modeling their foreign key graph structure. RelDiff combines a joint graph-conditioned diffusion process across all tables for attribute synthesis, and a $2K+$SBM graph generator based on the Stochastic Block Model for structure generation. The decomposition of graph structure and relational attributes ensures both high fidelity and referential integrity, both of which are crucial aspects of synthetic relational database generation. Experiments on 11 benchmark datasets demonstrate that RelDiff consistently outperforms prior methods in producing realistic and coherent synthetic relational databases. Code is available at https://github.com/ValterH/RelDiff.

cs.LG↗

arXiv2024

Benchmarking the Fidelity and Utility of Synthetic Relational Data

Synthesizing relational data has started to receive more attention from researchers, practitioners, and industry. The task is more difficult than synthesizing a single table due to the added complexity of relationships between tables. For the same reason, benchmarking methods for synthesizing relational data introduces new challenges. Our work is motivated by a lack of an empirical evaluation of state-of-the-art methods and by gaps in the understanding of how such an evaluation should be done. We review related work on relational data synthesis, common benchmarking datasets, and approaches to measuring the fidelity and utility of synthetic data. We combine the best practices and a novel robust detection approach into a benchmarking tool and use it to compare six methods, including two commercial tools. While some methods are better than others, no method is able to synthesize a dataset that is indistinguishable from original data. For utility, we typically observe moderate correlation between real and synthetic data for both model predictive performance and feature importance.

cs.DB↗

arXiv2021

Predicting the Popularity of Games on Steam

The video game industry has seen rapid growth over the last decade. Thousands of video games are released and played by millions of people every year, creating a large community of players. Steam is a leading gaming platform and social networking site, which allows its users to purchase and store games. A by-product of Steam is a large database of information about games, players, and gaming behavior. In this paper, we take recent video games released on Steam and aim to discover the relation between game popularity and a game's features that can be acquired through Steam. We approach this task by predicting the popularity of Steam games in the early stages after their release and we use a Bayesian approach to understand the influence of a game's price, size, supported languages, release date, and genres on its player count. We implement several models and discover that a genre-based hierarchical approach achieves the best performance. We further analyze the model and interpret its coefficients, which indicate that games released at the beginning of the month and games of certain genres correlate with game popularity.

cs.LG↗

arXiv2020

GPU-based Parallel Computation Support for Stan

This paper details an extensible OpenCL framework that allows Stan to utilize heterogeneous compute devices. It includes GPU-optimized routines for the Cholesky decomposition, its derivative, other matrix algebra primitives and some commonly used likelihoods, with more additions planned for the near future. Stan users can now benefit from large speedups offered by GPUs with little effort and without changes to their existing Stan code. We demonstrate the practical utility of our work with two examples - logistic regression and Gaussian Process regression.

cs.MS↗

arXiv2019

bayes4psy -- an Open Source R Package for Bayesian Statistics in Psychology

Research in psychology generates interesting data sets and unique statistical modelling tasks. However, these tasks, while important, are often very specific, so appropriate statistical models and methods cannot be found in accessible Bayesian tools. As a result, the use of Bayesian methods is limited to those that have the technical and statistical fundamentals that are required for probabilistic programming. Such knowledge is not part of the typical psychology curriculum and is a difficult obstacle for psychology students and researchers to overcome. The goal of the bayes4psy package is to bridge this gap and offer a collection of models and methods to be used for data analysis that arises from psychology experiments and as a teaching tool for Bayesian statistics in psychology. The package contains Bayesian t-test and bootstrapping and models for analyzing reaction times, success rates, and colors. It also provides all the diagnostic, analytic and visualization tools for the modern Bayesian data analysis workflow.

stat.AP↗