Prediksi Kualitas Laporan Keuangan Kementerian Negara/Lembaga Menggunakan Machine Learning

Teguh Puspandoyo (Direktorat Jenderal Perbendaharaan, Kementerian Keuangan, Jakarta)
Widha Adinata (Direktorat Jenderal Perbendaharaan, Kementerian Keuangan, Jakarta)
Laode Atthar Alfikry (Direktorat Jenderal Perbendaharaan, Kementerian Keuangan, Jakarta)
Tyas Ayu Prasanti (Direktorat Jenderal Perbendaharaan, Kementerian Keuangan, Jakarta)
Silvi Yuniar (Direktorat Jenderal Perbendaharaan, Kementerian Keuangan, Jakarta)

Jurnal Manajemen Perbendaharaan
Vol 3 No 1 (2022): Jurnal Manajemen Perbendaharaan

Quality financial reports are one of the tools in realizing government accountability. However, not all State Ministries/Agencies yet have received the best opinion. Based on previous research, several factors have a significant influence on the quality of financial reports. Therefore, this study aims to analyze the influencing factors and also make predictions and classification of quality of state ministries/agencies of financial reports using machine learning techniques. The sample used in this study was 256 state ministries/agencies from 2018-2020. The results show that cash fraud, findings, follow-up to BPK's recommendations, the money supply/additional money supply, and realization have a strong influence in determining quality of financial reports. In addition, with the Support Vector Machine model, better classification predictions are obtained for quality of state ministries/agencies of financial reports.