Open AccessPENGEMBANGAN EARLY WARNING SYSTEM UNTUK DELISTING SAHAM SYARIAH MENGGUNAKAN SUPPORT VECTOR MACHINE (SVMs)
Author(s) -
Zulnani Tinggi,
Sakum
Publication year - 2020
Publication title -
jurnal ekonomi syariah pelita bangsa/jurnal ekonomi syariah pelita bangsa
Language(s) - English
Resource type - Journals
eISSN - 2745-7621
pISSN - 2528-0244
DOI - 10.37366/jespb.v5i02.117
Subject(s) - support vector machine , nonprobability sampling , leverage (statistics) , debt to equity ratio , business , sample (material) , debt , stock exchange , population , data mining , econometrics , computer science , accounting , artificial intelligence , finance , economics , medicine , chemistry , environmental health , chromatography
This study aim to produce Early Warning System in predicting the occurrence of delisting in Islamic stocks by using Support Vector Machines (SVM). The sample used in this research are companies listed on the Indonesian Syariah Stock Index (ISSI) for the period of 2012 - 2018. With the variables used in this research: Turn Over Asset, Long Term Debt, Interest Coverage, Debt to Equity, Quick Ratio, ROA, ROE Leverage, Current Ratio, ROIC. The population of this study is 335 Islamic stocks registered with ISSI. There are 102 companies which consists of listed and delisted companies from sharia shares as comparison for the sample data. The Method applied in this study is Purposive Sampling for The sampling technique. From the result found that accuracy rate of the best SVM models is SVM 4 models with 100% accuracy