Data squashing as preprocessing in association rule mining

Data squashing is a well-known preprocessing method in Machine Learning that enables construction of smaller datasets from the original ones and provides approximately the same results of data analysis as the original. The paper proposes a new data squashing method for Association Rule Mining based...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2022 IEEE Symposium Series on Computational Intelligence (SSCI) s. 1720 - 1725
Hlavní autoři: Fister, Iztok, Novak, Damijan, Verber, Domen
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 04.12.2022
Témata:
association rule mining
Computational intelligence
Data analysis
Data mining
data squashing
Euclidean distance
evolutionary algorithms
Machine learning
Measurement
preprocessing
Testing
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Data squashing is a well-known preprocessing method in Machine Learning that enables construction of smaller datasets from the original ones and provides approximately the same results of data analysis as the original. The paper proposes a new data squashing method for Association Rule Mining based on the Cosine similarity and Euclidean distance similarity. The method was applied to three datasets from the UCI Machine Learning repository. The results showed that the proposed data squashing method is effective, scalable, and easy to use, and therefore represents a huge potential for use in practice.
DOI:10.1109/SSCI51031.2022.10022240