Electronic Resource
Perbandingan Model Prediksi Delisting Pada Saham Syariah Berbasis Informasi Keuangan Studi Perbandingan Regresi Logistik, Neural, Networks Dan Support Vector Machine Pada Indeks Saham Syariah Indonesia
"The aim of this study is to find the best model in predicting the occurrence of
delisting in Islamic stocks by comparing three types of models, Logistic
Regression, Artificial Neural Network (ANNs) and Support Vector Machines
(SVM) in the companies listed on the Indonesian Syariah Stock Index (ISSI) in the
period 2012 - 2018. With the variables ROA, ROE Leverage, Debt to Equity,
Quick Ratio, Current Ratio, ROIC, Turn Over Asset, Long Term Debt, Interest
Coverage. The population of this study is 335 Islamic stocks registered with ISSI
in the period 2012 - 2018. The samples in this research are 102 companies which
consists of listed and delisted companies from sharia shares as comparison. The
sampling method in this study is Purposive Sampling. Comparison of the analytical
model used in this study are logistic regression, Artificial neural Networks and
Support vector Machines. Logistic Regression result is 93.85% with Normit
model, the ANNs models is the MLP 10-16-16 with 98.04% accuracy rate and
the best SVM models is SVM 4 models with 100% accuracy"
No other version available