aiqbench analysis
This paper proposes MOEA-SSL, a multi-objective evolutionary method for semi-supervised tabular classification that trades predictive accuracy against the number of features used. Its reported gains are reduced feature usage relative to the prior EA-SSL method, with competitive accuracy against lightweight SSL baselines on three smart-city datasets at 1%, 5%, and 10% labeled data. For practitioners, the useful takeaway is the Pareto front of lower-feature alternatives, though the evidence comes from a small set of datasets and a conference-level evaluation.
Multi-Objective Evolutionary Semi-Supervised Learning for Resource-Aware Smart City Classification
arXiv:2610.27703v1 Announce Type: new Abstract: Smart-city applications often combine scarce labeled data with abundant unlabeled observations. Semi-supervised learning (SSL) exploits both labeled and unlabeled data to improve classification under limited supervision. This article presents MOEA-SSL, a multi-objective evolutionary approach for lightweight semi-supervised tabular classification that jointly optimizes predictive quality and feature usage. Experimental evaluation on three Smart City classification datasets, considering 1\%, 5\%, and 10\% labeled data, shows that MOEA-SSL achieves competitive predictive performance compared with established lightweight SSL methods while significantly reducing feature usage relative to EA-SSL in all evaluated conditions. The computed Pareto fronts also provide lower-feature-use alternatives with limited degradation in classification performance.
Abstract:Smart-city applications often combine scarce labeled data with abundant unlabeled observations. Semi-supervised learning (SSL) exploits both labeled and unlabeled data to improve classification under limited supervision. This article presents MOEA-SSL, a multi-objective evolutionary approach for lightweight semi-supervised tabular classification that jointly optimizes predictive quality and feature usage. Experimental evaluation on three Smart City classification datasets, considering 1\%, 5\%, and 10\% labeled data, shows that MOEA-SSL achieves competitive predictive performance compared with established lightweight SSL methods while significantly reducing feature usage relative to EA-SSL in all evaluated conditions. The computed Pareto fronts also provide lower-feature-use alternatives with limited degradation in classification performance.
ICSC-CITIES 2026. IX Ibero-American Congress of Smart Cities
Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
arXiv:2610.27703 [cs.LG]
(or arXiv:2610.27703v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.27703 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 208 words