Publisher

FKPTFKPT

Journal Bulletin of Information Technology (BIT)Bulletin of Information Technology (BIT) Abstract

The National Health Insurance (JKN) program, managed by BPJS Kesehatan, has experienced a significant increase in healthcare service costs, particularly at Advanced Referral Healthcare Facilities (FKRTL). This study aims to compare the forecasting accuracy of ARIMA and Long Short-Term Memory (LSTM) methods in predicting healthcare service costs in FKRTL Bogor from January 2014 to October 2024. The data, sourced from BPJS Kesehatan Branch Bogor, were analyzed using time series approaches. Model evaluation was conducted using Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). Results show that for 80% of training data, LSTM produced a MAPE of 8.85% and RMSE of IDR 6.98 billion, slightly outperforming ARIMA (0,1,1) with MAPE of 10.28% and RMSE of IDR 6.67 billion. For the 20% testing data, LSTM demonstrated significantly better accuracy, with an MAPE of 12.97% and RMSE of IDR 15.52 billion, compared to ARIMAs MAPE of 24.22% and RMSE of IDR 30.76 billion. Therefore, LSTM is considered more effective for short- to medium-term forecasting of JKN healthcare costs, particularly when dealing with complex and non-linear patterns.

Conclusion

The study concludes that LSTM outperforms ARIMA in forecasting healthcare service costs at FKRTL Bogor, particularly in the testing phase.LSTM demonstrated superior accuracy with lower MAPE and RMSE values compared to ARIMA.This suggests that LSTM is a more effective method for short-to-medium-term forecasting of JKN healthcare costs, especially when dealing with complex and non-linear patterns in the data.

Future Research

Penelitian lebih lanjut dapat dilakukan untuk menguji efektivitas model LSTM dengan menggabungkannya dengan metode lain, seperti Convolutional Neural Networks (CNN), untuk meningkatkan akurasi prediksi biaya layanan kesehatan. Selain itu, studi dapat diperluas dengan memasukkan variabel eksternal, seperti data demografi, tingkat inflasi, dan kejadian penyakit menular, ke dalam model LSTM untuk melihat apakah faktor-faktor ini dapat meningkatkan kemampuan prediksi. Terakhir, penelitian dapat difokuskan pada pengembangan model yang dapat memprediksi biaya layanan kesehatan berdasarkan jenis penyakit atau prosedur medis tertentu, sehingga memungkinkan BPJS Kesehatan untuk mengalokasikan sumber daya secara lebih efisien dan efektif, serta merencanakan strategi intervensi kesehatan yang lebih tepat sasaran.

References

1. Hybrid ARIMA-LSTM for COVID-19 forecasting: a comparative AI modeling study [PeerJ]. hybrid arima lstm... peerj.com/articles/cs-3195Hybrid ARIMA LSTM for COVID 19 forecasting a comparative AI modeling study PeerJ hybrid arima lstm peerj articles cs 3195
2. 0. scitepress publication details scitepress.org/Link.aspx?doi=10.5220/00123626000036540 scitepress publication details scitepress Link aspx doi 10 5220 0012362600003654
3. Comparative study of long short-term memory (LSTM), bidirectional LSTM, and traditional machine learning... doi.org/10.1177/01445987241269496Comparative study of long short term memory LSTM bidirectional LSTM and traditional machine learning doi 10 1177 01445987241269496

Metadata

File size	740.91 KB
Pages	9
DMCA	Report