Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mendez, David, Martin-Maroto, Fernando, de Polavieja, Gonzalo G.
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2605.22155
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911704325029888
author Mendez, David
Martin-Maroto, Fernando
de Polavieja, Gonzalo G.
author_facet Mendez, David
Martin-Maroto, Fernando
de Polavieja, Gonzalo G.
contents Symbolic methods are generally not considered competitive with strong modern learners on realistic supervised tasks. We evaluate Algebraic Machine Learning (AML), a framework that learns through subdirect decomposition of algebraic structure rather than numerical optimization, against standard baselines on image and tabular classification across varying training-set sizes. We find that AML trained only on training data without using validation or cross-validation outperforms a family of cross-validated baseline methods including CNNs on small to medium image datasets (50--2000 training examples). On tabular datasets in the same size range, XGBoost is overall the best performing method, but AML is nonetheless comparable to methods incorporating task-specific biases such as LightGBM and random forests. AML achieves this competitive performance across two very different types of datasets using a generic algebraic inductive bias, rather than the modality-specific biases built into standard baselines like CNNs for images or XGBoost for tabular data, and requires no cross validation because it has no task-dependent hyperparameters to tune.
format Preprint
id arxiv_https___arxiv_org_abs_2605_22155
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Algebraic Machine Learning for Small-to-Medium Datasets Is Competitive against Strong Standard Baselines
Mendez, David
Martin-Maroto, Fernando
de Polavieja, Gonzalo G.
Machine Learning
I.2.6
Symbolic methods are generally not considered competitive with strong modern learners on realistic supervised tasks. We evaluate Algebraic Machine Learning (AML), a framework that learns through subdirect decomposition of algebraic structure rather than numerical optimization, against standard baselines on image and tabular classification across varying training-set sizes. We find that AML trained only on training data without using validation or cross-validation outperforms a family of cross-validated baseline methods including CNNs on small to medium image datasets (50--2000 training examples). On tabular datasets in the same size range, XGBoost is overall the best performing method, but AML is nonetheless comparable to methods incorporating task-specific biases such as LightGBM and random forests. AML achieves this competitive performance across two very different types of datasets using a generic algebraic inductive bias, rather than the modality-specific biases built into standard baselines like CNNs for images or XGBoost for tabular data, and requires no cross validation because it has no task-dependent hyperparameters to tune.
title Algebraic Machine Learning for Small-to-Medium Datasets Is Competitive against Strong Standard Baselines
topic Machine Learning
I.2.6
url https://arxiv.org/abs/2605.22155