Magnitude Distance: A Geometric Measure of Dataset Similarity

Sahel Torkamani, Henry Gouk, Rik Sarkar
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:121792-121810, 2026.

Abstract

Quantifying the distance between datasets is a fundamental question in mathematics and machine learning. We propose magnitude distance, a novel distance metric defined on finite datasets using the notion of the magnitude of a metric space. The proposed distance incorporates a tunable scaling parameter, $t$, that controls the sensitivity to global structure (small $t$) and finer details (large $t$). We prove several theoretical properties of magnitude distance, including its limiting behavior across scales and conditions under which it satisfies key metric properties. In contrast to classical distances, we show that magnitude distance remains discriminative in high-dimensional settings when the scale is appropriately tuned. We further demonstrate how magnitude distance can be used as a training objective for push-forward generative models. Our experimental results support our theoretical analysis and demonstrate that magnitude distance provides meaningful signals, comparable to established distance-based generative approaches.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-torkamani26a, title = {Magnitude Distance: A Geometric Measure of Dataset Similarity}, author = {Torkamani, Sahel and Gouk, Henry and Sarkar, Rik}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {121792--121810}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/torkamani26a/torkamani26a.pdf}, url = {https://proceedings.mlr.press/v306/torkamani26a.html}, abstract = {Quantifying the distance between datasets is a fundamental question in mathematics and machine learning. We propose magnitude distance, a novel distance metric defined on finite datasets using the notion of the magnitude of a metric space. The proposed distance incorporates a tunable scaling parameter, $t$, that controls the sensitivity to global structure (small $t$) and finer details (large $t$). We prove several theoretical properties of magnitude distance, including its limiting behavior across scales and conditions under which it satisfies key metric properties. In contrast to classical distances, we show that magnitude distance remains discriminative in high-dimensional settings when the scale is appropriately tuned. We further demonstrate how magnitude distance can be used as a training objective for push-forward generative models. Our experimental results support our theoretical analysis and demonstrate that magnitude distance provides meaningful signals, comparable to established distance-based generative approaches.} }
Endnote
%0 Conference Paper %T Magnitude Distance: A Geometric Measure of Dataset Similarity %A Sahel Torkamani %A Henry Gouk %A Rik Sarkar %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-torkamani26a %I PMLR %P 121792--121810 %U https://proceedings.mlr.press/v306/torkamani26a.html %V 306 %X Quantifying the distance between datasets is a fundamental question in mathematics and machine learning. We propose magnitude distance, a novel distance metric defined on finite datasets using the notion of the magnitude of a metric space. The proposed distance incorporates a tunable scaling parameter, $t$, that controls the sensitivity to global structure (small $t$) and finer details (large $t$). We prove several theoretical properties of magnitude distance, including its limiting behavior across scales and conditions under which it satisfies key metric properties. In contrast to classical distances, we show that magnitude distance remains discriminative in high-dimensional settings when the scale is appropriately tuned. We further demonstrate how magnitude distance can be used as a training objective for push-forward generative models. Our experimental results support our theoretical analysis and demonstrate that magnitude distance provides meaningful signals, comparable to established distance-based generative approaches.
APA
Torkamani, S., Gouk, H. & Sarkar, R.. (2026). Magnitude Distance: A Geometric Measure of Dataset Similarity. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:121792-121810 Available from https://proceedings.mlr.press/v306/torkamani26a.html.

Related Material