Naslov
Supervised and Unsupervised Machine Learning Approaches on Class Imbalanced Data
(Supervised and Unsupervised Machine Learning Approaches on Class Imbalanced Data)

Autori
Ugarković, Alen ; Oreški, Dijana

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of International Conference on Smart Systems and Technologies (SST 2022) / Nyarko, Emmanuel Karlo ; Matić, Tomislav ; Cupec, Robert ; Vranješ, Mario - Osijek : Fakultet elektrotehnike, računarstva i informacijskih tehnologija Sveučilišta Josipa Jurja Strossmayera u Osijeku, 2022, 149-152

ISBN
978-1-6654-8214-1

Skup
International Conference on Smart Systems and Technologies 2022 (SST 2022)

Mjesto i datum
Osijek, Hrvatska, 19.10.2022. - 21.10.2022

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
class imbalance ; cluster analysis ; decision tree ; machine learning.

Sažetak
Huge amounts of data are stored digitally every day. This data has various characteristics. Class imbalance is one of the characteristics that has the effect of machine learning algorithms performance and this problem is receiving attention among academia and industry. Class imbalance occurs when the number of instances in one class is significantly different than the number of instances in the other class (in binary classification). In this paper, we are combining supervised and unsupervised machine learning approaches on one imbalanced dataset from a publicly available repository. Unsupervised machine learning approach of cluster analysis is applied on the most significant variables discovered by sensitivity analysis on predictive models developed by decision tree. Our results indicated a hybrid approach of decision tree and cluster analysis as a promising tool to work with imbalanced data.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti

POVEZANOST RADA