Performance evaluation of machine learning approaches to optimize it service management

Mohammad Agus Prihandono, Riri Fitri Sari

Research output: Contribution to journalArticlepeer-review


IT service management experienced a variety of incidents that can be categorized as incident attributes. Incident attributes can be considered as an important IT cycle relationship in an organization. These attributes of the incident might improve the event response of an IT cycle in the occurrence of disruption and to identify the lowest possible impact in an organization. In this research, we used support vector machine (SVM) classifier, using both genetic algorithm (GA) and particle swarm optimization (PSO) optimization technique. The aim is to gain better accuracy. Furthermore, rules are extracted to recognize the weight of IT incident. Our experiment result shows that SVM with optimization algorithms using particle swarm optimization is better than using genetic algorithm.

Original languageEnglish
Pages (from-to)3518-3525
Number of pages8
JournalInternational Journal of Advanced Science and Technology
Issue number7 Special Issue
Publication statusPublished - 14 Apr 2020


  • Classification
  • IT service management
  • Machine learning
  • Optimization


Dive into the research topics of 'Performance evaluation of machine learning approaches to optimize it service management'. Together they form a unique fingerprint.

Cite this