VSE knjižnice (vzajemna bibliografsko-kataložna baza podatkov COBIB.SI)
  • CMAC: Clustering class association rules to form a compact and meaningful associative classifier [Elektronski vir]
    Jamolbek Maqsudovich, Mattiev ; Kavšek, Branko
    Huge amounts of data are being collected and analyzed nowadays. By using the popular rule-learning algorithms, the number of rules discovered on those big datasets can easily exceed thousands of ... rules. To produce compact and accurate classifiers, such rules have to be grouped and pruned, so that only a reasonable number of them are presented to the end user for inspection and further analysis. To solve this problem researchers have proposed several associative classification approaches that combine two important data mining techniques, namely, classification and association rule mining. In this paper, we propose a new method that is able to reduce the number of class association rules produced by classical class association rule classifiers, while maintaining an accurate classification model that is comparable to the ones generated by state-of-the-art classification algorithms. More precisely, we propose a new associative classifier – CMAC, that uses agglomerative hierarchical clustering as a post-processing step to reduce the number of its rules. Experimental results performed on selected datasets from the UCI ML repository show that CMAC is able to learn classifiers containing significantly less rules than state-of-the-art rule learning algorithms on datasets with larger number of examples. On the other hand, classification accuracy of the CMAC classifier is not significantly different from state-of-the-art rule-learners on most of the datasets. We can thus conclude that CMAC is able to learn compact (and meaningful) classifiers from “bigger” datasets, retaining an accuracy comparable to state-of-the-art rule learning algorithms.
    Vir: Machine Learning, Optimization, and Data Science [Elektronski vir] : 6th International Conference, LOD 2020 Siena, Italy, July 19–23, 2020 : revised selected papers. Part I (str. 372-384)
    Vrsta gradiva - prispevek na konferenci ; neleposlovje za odrasle
    Leto - 2020
    Jezik - angleški
    COBISS.SI-ID - 94833155

    Povezava(-e):

    https://link.springer.com/chapter/10.1007%2F978-3-030-64583-0_34
    https://doi.org/10.1007/978-3-030-64583-0_34
    DOI

Vnos na polico