arXiv subjects

Catherine B. Hurley

Publications and source records attributed to Catherine B. Hurley.

3 recordsLinked to original sources

arXiv2025

A Tidy Data Structure and Visualisations for Multiple Variable Correlations and Other Pairwise Scores

We provide a pipeline for calculating, managing and visualising correlations and other pairwise association scores for numerical and categorical data. We present a uniform interface for calculating a plethora of pairwise scores and propose a tidy data structure for organising the results. We also provide new visualisations which simultaneously show multiple and/or grouped pairwise scores. The visualisations are far richer than a traditional heatmap of correlation scores, as they help identify relationships with categorical variables, numeric variable pairs with non-linear associations or those which exhibit Simpson's paradox. These methods are available in our R package bullseye.

stat.CO↗

arXiv2021

Interactive slice visualization for exploring machine learning models

Machine learning models fit complex algorithms to arbitrarily large datasets. These algorithms are well-known to be high on performance and low on interpretability. We use interactive visualization of slices of predictor space to address the interpretability deficit; in effect opening up the black-box of machine learning algorithms, for the purpose of interrogating, explaining, validating and comparing model fits. Slices are specified directly through interaction, or using various touring algorithms designed to visit high-occupancy sections or regions where the model fits have interesting properties. The methods presented here are implemented in the R package \pkg{condvis2}.

stat.ML↗

arXiv2016

Conditional Visualization for Statistical Models: An Introduction to the condvis Package in R

The condvis package is for interactive visualization of sections in data space, showing fitted models on the section, and observed data near the section. The primary goal is the interpretation of complex models, and showing how the observed data support the fitted model. There is a video accompaniment to this paper available at https://www.youtube.com/watch?v=rKFq7xwgdX0. This is a preprint version of an article to appear in the Journal of Statistical Software.

stat.OT↗