PREDICTING TIMELINESS OF CORPORATE FINANCIAL REPORTING BASED ON SUPPORT VECTOR MACHINE

Erna Hayati (Faculty of Economics, Universitas Islam Lamongan)
Diah Ayu Novitasari (Faculty of Economics, Universitas Islam Lamongan)
Rivatul Ridho Elvierayani (Faculty of Economics, Universitas Islam Lamongan)
Tarisa Ayu Firnanda (Faculty of Economics, Universitas Islam Lamongan)

Jurnal Ekonomi
Vol. 11 No. 03 (2022): Jurnal Ekonomi, 2022 Periode Desember

The main purpose of this study is to predict the timeliness of the corporate financial reports with Support Vector Machine (SVM) and comparing the performance of SVM with Linear Discriminant Analysis (LDA). 30 samples of Food and Beverages companies on the Indonesia Stock Exchange (IDX) were used in this study. This study concludes that SVM has the best performance in predicting the timeliness of financial reporting compared to LDA. The best variables that can significantly distinguish companies that are on time and not are Return on Assets (ROA) and Debt to Asset Ratio (DAR). Meanwhile, the other two variables, namely Current Ratio (CR) and Company Size, have no significant effect. These results are expected to be a reference for users of financial information in making decisions