Claim Missing Document
Check
Articles

Found 12 Documents
Search

PENGGUNAAN METODE BORNHUETTER-FERGUSON UNTUK ESTIMASI CADANGAN KLAIM Riaman Riaman; Betty Subartini; Kankan Parmikanti
Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika Vol. 4 No. 2 (2023): Jurnal Lebesgue : Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistik
Publisher : LPPM Universitas Bina Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46306/lb.v4i2.366

Abstract

Health insurance company have to determine claim reserve that’s suitable with the existing condition. There is three party that’s involved in the health insurance management, namely the policy holder, Admedika as the third party administration, and also the insurance company itself as the (insurer). When the policy holder obtained treatments whose financing is done through a health insurance, then the health insurance company have the obligation to finish the financial matters. Delays in payments from insurance companies to health facilities are caused, among others, by the administrative process. Thus, every claim submitted by the insured party to the insurance company will be settled in stages to the health facilities. The data presented from these conditions form a triangle matrix (run-off triangle) which then becomes the basis for estimating the amount of IBNR claims reserves. The Bornhuetter-Ferguson (BF) method involves the amount of premium that has become income for the company and calculates the Ultimate Claim value in estimating the amount of claim reserves. This method is the result of the development of the previous method, Chain-Ladder (CL), which only relies on historical data on claim payments. Premium calculations need to be involved in health insurance, because the insurance period is short, which is only one year. Insurance companies haven't had time to turn around the money to invest, so payment of claims will depend more on the premium that becomes income for the company (earned premium). The estimated claim reserve value is more suitable and robust than the CL method. Estimated claim reserves that occur in the 2nd event period amount to IDR50,658,714 with an estimated interval for the 2nd event period between IDR10,215,477 and IDR91,101,950
Analisis Perbandingan Hasil Peramalan Harga Saham Menggunakan Model Autoregresive Integrated Moving Average dan Long Short Term Memory Luki Setiawan; Dwi Susanti; Riaman Riaman
Jurnal Matematika Integratif Vol 19, No 2: Oktober 2023
Publisher : Department of Matematics, Universitas Padjadjaran

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24198/jmi.v19.n2.42164.223-234

Abstract

Saham menjadi salah satu instrumen investasi yang populer di tengah masyarakat modern. Saham berpotensi memberikan keuntungan yang besar namun juga memiliki risiko yang besar, oleh sebab itu dibutuhkan peramalan harga saham untuk menghadapi risiko dalam berinvestasi saham. Data harga saham termasuk ke dalam data deret waktu sehingga diperlukan analisis deret waktu dalam meramalkannya. Terdapat dua model populer dalam meramalkan data deret waktu yaitu Model Autoregressive Integrated Moving Average (ARIMA) dan Model Long Short Term Memory (LSTM). Tujuan dalam penelitian ini adalah untuk menemukan model ARIMA terbaik dan kombinasi hyperparameter model LSTM terbaik, serta membandingkan akurasi hasil peramalan kedua model tersebut untuk memperoleh model yang terbaik dalam meramalkan harga saham terpilih. Metode Maximum Likelihood Estimation digunakan dalam mengestimasi parameter model ARIMA dan Metode Trial and Error digunakan dalam menentukan kombinasi hyperparameter model LSTM. Data yang digunakan adalah data harga penutupan saham BBCA, BBTN, dan BMRI selama 1 tahun (1 April 2021 – 31 Maret 2022). Hasil penelitian menunjukkan bahwa model LSTM merupakan model terbaik dalam meramalkan data harga saham BBCA, sementara itu model ARIMA (1,1,0) merupakan model terbaik dalam meramalkan data harga saham BBTN dan BMRI. Seluruh hasil peramalan dengan menggunakan model terbaik untuk masing-masing saham, masuk ke dalam kriteria peramalan yang sangat akurat karena memiliki nilai MAPE <10%.