Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Hybrid Intelligent Parsimony Search in Small High-Dimensional Datasets
Ist Teil von
Hybrid Artificial Intelligent Systems, p.384-396
Ort / Verlag
Cham: Springer Nature Switzerland
Link zum Volltext
Quelle
Alma/SFX Local Collection
Beschreibungen/Notizen
The search for machine learning models that generalize well with small high-dimensional datasets is a current challenge. This paper shows a specific hybrid methodology for this kind of problems combining HYB-PARSIMONY and Bayesian Optimization. The methodology proposes to use HYB-PARSIMONY with different random seeds and select those features that had the highest mean probability. Subsequently, with these features, a hyperparameter adjustment is performed with Bayesian Optimization. The results show that the methodology substantially improves the degree of generalization and parsimony of the obtained models compared to previous methods.