Abstract Latent feature representation methods play an important role in dimension reduction and statistical modeling of high-dimensional complex data objects. However, existing approaches to evaluate these methods have two main drawbacks: they often rely on aggregate measures of information loss, such as average or total loss, which can mask variation across observations, and they typically do so using training rather than generalization error.
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Giles Hooker
When AI Transparency Backfires
The following article is written by Fei Huang, associate professor of risk and actuarial studies at UNSW Business School, and Giles Hooker, professor of statistics and data science at Wharton. Senior leaders across industries face growing pressure to demonstrate that their AI systems are transparent, fair, and well governed. Regulators want explanations. Boards commission dashboards. Customers expect to understand how decisions that affect them are reached.
The importance of accounting for spatial heterogeneity in studies of plant competition and coexistence
CONFLICT OF INTEREST STATEMENT The authors have no conflicts to declare. PEER REVIEW The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/1365-2745.70221. DATA AVAILABILITY STATEMENT No original data were collected for this study. The empirically derived parameters and all code associated with this study are available in FigShare: https://doi.org/10.6084/m9.figshare.28924001 (Schiffer et al., 2025).
Computer Science > Machine Learning
[Submitted on 6 Feb 2025 (v1), last revised 27 Oct 2026 (this version, v2)] View PDF HTML (experimental) Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) Cite as: arXiv:2502.04309 [cs.LG] (or arXiv:2502.04309v2 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2502.04309
Why You Should Not Trust Interpretations in Machine Learning: Adversarial Attacks on Partial Dependence Plots
[Submitted on 29 Apr 2024 ( v1 ), last revised 1 May 2024 (this version, v2)] Title:Why You Should Not Trust Interpretations in Machine Learning: Adversarial Attacks on Partial Dependence Plots View a PDF of the paper titled Why You Should Not Trust Interpretations in Machine Learning: Adversarial Attacks on Partial Dependence Plots, by Xi Xin and 2 other authors View PDF HTML (experimental) Abstract:The adoption of artificial intelligence (AI) across industries has led to the widespread use...
[2401.15800] Statistical Significance of Feature Importance Rankings
arXiv:2401.15800 (stat) [Submitted on 28 Jan 2024 (v1), last revised 29 Jan 2025 (this version, v3)] View PDF HTML (experimental) Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG) Cite as: arXiv:2401.15800 [stat.ML] (or arXiv:2401.15800v3 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2401.15800 Submission history From: Jeremy Goldwasser [ view email] [v1] Sun, 28 Jan 2024 23:14:51 UTC (180 KB) [v2] Mon, 3 Jun 2024 00:49:43 UTC (247 KB) [v3] Wed, 29 Jan 2025 00:49:57...
[2310.07672] Stabilizing Estimates of Shapley Values with Control Variates
arXiv:2310.07672 (stat) [Submitted on 11 Oct 2023 (v1), last revised 10 Apr 2024 (this version, v3)] View PDF HTML (experimental) Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG) Cite as: arXiv:2310.07672 [stat.ML] (or arXiv:2310.07672v3 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2310.07672 Submission history From: Jeremy Goldwasser [ view email] [v1] Wed, 11 Oct 2023 17:18:51 UTC (518 KB) [v2] Fri, 10 Nov 2023 01:28:50 UTC (542 KB) [v3] Wed, 10 Apr 2024 00:35:36...
A practical guide to selecting models for exploration, inference, and prediction in ecology
Model selection is the Black Hole of Statistics. A. Ronald Gallant, Liberal Arts Professor of Economics, Pennsylvania State University, and Henry A. Latane Distinguished Professor (emeritus) of Economics, UNC-Chapel Hill (personal communication) Introduction In 2014, Ecology ran a productive and contentious Forum titled "P values, hypothesis testing, and model selection: it's déjà vu all over again" (Ellison et al. 2014).
Stop Permuting Features
Homer Simpsons is known to be a simple person, who makes silly deeds from time to time. Here he decided to cook bread, bacon, and eggs using the gas stove as it was a bonfire; the tool which isn’t suitable for such purposes. We found it funny to watch, but at the same time, we tend to do similar stuff in data science. Don’t be like Homer, use proper tools.
Racism, Redcards, and Rabbit Holes
Statisticians have warned for a long time that inexpert modeling can produce distorted statistical results. The replication crisis in science has often been attributed to poor statistical practice with part of the cure being to employ more statisticians. But would this really help? An interesting study by Silberzahn et. al. was posted earlier this year that looked at how consistent statisticians themselves were in their analysis.
Online Dating and the Statistical Dark Arts
One of the darkest statistical arts lies in choosing the model to use when analyzing your experimental data. A statistical model both represents your understanding of the experiment and enables you to test the strength of evidence supporting your conclusions.
Online Dating and Other Encounters with Confounders
One of the darkest statistical arts lies in choosing the model to use when analyzing your experimental data. A statistical model both represents your understanding of the experiment and enables you to test the strength of evidence supporting your conclusions.
Show More
loading
Actions
Is this you?
As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.Get in touch with Giles
Contact Giles, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack