Utilisasi algoritma Backpropagation Artificial Neural Network pada sistem prediksi target Pendapatan Pajak Daerah

Laksana, Satria (2024) Utilisasi algoritma Backpropagation Artificial Neural Network pada sistem prediksi target Pendapatan Pajak Daerah. Sarjana thesis, UIN Sunan Gunung Djati Bandung.

Preview	Text (COVER) 1_cover.pdf Download (45kB) | Preview
Preview	Text (ABSTRAK) 2_abstrak.pdf Download (87kB) | Preview
Preview	Text (DAFTAR ISI) 3_daftarisi.pdf Download (88kB) | Preview
Preview	Text (BAB I) 4_bab1.pdf Download (127kB) | Preview
	Text (BAB II) 5_bab2.pdf Restricted to Registered users only Download (466kB) | Request a copy
	Text (BAB III) 6_bab3.pdf Restricted to Registered users only Download (98kB) | Request a copy
	Text (BAB IV) 7_bab4.pdf Restricted to Registered users only Download (5MB) | Request a copy
	Text (BAB V) 8_bab5.pdf Restricted to Registered users only Download (28kB) | Request a copy
	Text (DAFTAR PUSTAKA) 9_daftarpustaka.pdf Restricted to Registered users only Download (87kB) | Request a copy

Abstract

INDONESIA: Pajak daerah memiliki peran penting dalam mendukung keuangan daerah dan pembangunan di tingkat lokal. Namun, metode yang digunakan dalam menetapkan target pendapatan pajak daerah, seperti pajak kendaraan bermotor, sering kali hanya mengandalkan pendekatan sederhana, sehingga kurang optimal dalam memanfaatkan data historis. Penelitian ini mengusulkan solusi dengan memanfaatkan teknologi machine learning, khususnya algoritma Artificial Neural Network (ANN) dengan metode backpropagation, untuk meningkatkan akurasi prediksi target pendapatan pajak kendaraan bermotor di wilayah Jawa Barat. Dengan mengembangkan Sistem Prediksi Target Pendapatan Pajak Daerah menggunakan BP-ANN dan menggunakan data pajak kendaraan bermotor dari Jawa Barat dari tahun 2013 hingga 2021, penelitian ini bertujuan untuk memberikan kontribusi dalam pengembangan aplikasi machine learning secara sistematis dalam prediksi pendapatan pajak. Metodologi AI Project Life Cycle yang diusulkan memandu penelitian ini mulai dari perencanaan hingga implementasi, mendorong pendekatan yang terstruktur. Hasil penelitian menunjukkan bahwa model BP-ANN secara signifikan mengungguli performa baseline yang sebelumnya ditetapkan. RMSE pada model BP-ANN sebesar 34.99%, sedangkan pada baseline sebesar 44.70%. Lebih lanjut, Explained Variance Score pada model BP-ANN mencapai 32.60%, sementara pada baseline hanya 12.17%. ENGLISH: Local taxes play a crucial role in supporting regional finances and local-level development. However, the methods used to determine local tax revenue targets, such as motor vehicle taxes, often rely on simple approaches like comparing the number of vehicles between previous years, leading to suboptimal utilization of historical data. This research proposes a solution by leveraging machine learning technology, specifically the Artificial Neural Network (ANN) algorithm with backpropagation method, to enhance the accuracy of predicting motor vehicle tax revenue targets in the West Java region. By developing a Local Tax Revenue Target Prediction System using BP-ANN and utilizing motor vehicle tax data from West Java from 2013 to 2021, this study aims to contribute to the systematic development of machine learning applications in tax revenue prediction. The proposed AI Project Life Cycle methodology guides this research from planning to implementation, promoting a structured approach. Research results demonstrate that the BP-ANN model significantly outperforms the previously established baseline performance. The RMSE of the BP-ANN model is 34.99%, whereas the baseline is 44.70%. Furthermore, the Explained Variance Score of the BP-ANN model reaches 32.60%, compared to only 12.17% in the baseline.

Actions (login required)

View Item