ESTiG - Publicações em Proceedings Indexadas à WoS/Scopus
URI permanente para esta coleção:
Navegar
Percorrer ESTiG - Publicações em Proceedings Indexadas à WoS/Scopus por Domínios Científicos e Tecnológicos (FOS) "Ciências Naturais::Matemáticas"
A mostrar 1 - 2 de 2
Resultados por página
Opções de ordenação
- Comparison between single and multi-objective clustering algorithms: mathE case studyPublication . Azevedo, Beatriz Flamia; Rocha, Ana Maria A.C.; Fernandes, Florbela P.; Pacheco, Maria F.; Pereira, Ana I.This paper compares the results obtained for four single clustering algorithms with a multi-objective clustering approach. For this, a dataset describing the student’s behavior within the Linear Algebra topic on the MathE e-learning platform is used. This dataset aids in understanding student performance and engagement in MathE to support the development of an intelligent system to tailor the platform’s resources to users’s needs. The four algorithms suggested two clusters as the optimal solution for the dataset. However, this binary categorization did not provide meaningful insights into the proposal of the MathE platform; that is, it did not provide a customized system according to individual needs. Thus, this study uses the multi-objective clustering algorithm, which results in a set of non-dominated solutions, providing decision-makers with a broader range of options to choose the solution that best meets their needs. The results demonstrate the main benefits of the proposed human-in-the-loop multi-objective approach since it provides several optimal solutions and allows the decision-maker to apply fundamental knowledge to define the most appropriate solution to the problem based on previous knowledge.
- Reconstruction of meteorological records with PCA-based analog ensemble methodsPublication . Breve, Murilo M.; Balsa, Carlos; Rufino, JoséThe Analog Ensemble (AnEn) method has been used to reconstruct missing data in time series with base on other correlated time series with full data. As the AnEn method benefits from the use of large volumes of data, there is a great interest in improving its efficiency. In this paper, the Principal Component Analysis (PCA) technique is combined with the classical AnEn method and a K-means cluster-based variant, within the context of reconstructing missing meteorological data at a particular station using information from neighboring stations. This combination allows to reduce the dimension of the number of predictor time series, while ensuring better accuracy and higher computational performance than the AnEn methods: it reduces prediction errors by up to 30% and achieves a computational speedup of up to 2x.