Ferra Yanuar, Ridha Fadhila Sani, Aidinil Zetra, Yudiantri Asdi, Fenni Kurnia Mutiya
The purpose of this study is to determine the estimation of the shape parameter (θ) of the Lomax distribution assuming the scale parameter (β) is known. The estimation of the parameters of the Lomax distribution in this study uses several Bayesian loss function methods, namely Bayesian Square Error Loss Function (SELF), Bayesian Entropy Loss Function (ELF), and Bayesian Precautionary Loss Function (PLF). The selected prior distribution is using the Gamma conjugate prior and Jeffrey’s prior as a non-informative prior. The data used in this study are Lomax-distributed generation data with sample sizes of 30, 100, 150, and 300 with different shape parameters (θ = 1.3 and 1.5). By using the criteria of the smallest AIC, AICc, and BIC values, this study proves that the Bayesian SELF with Jeffrey prior is the best Bayesian loss function compared to the other two methods. © 2026 Author(s).
Department of Mathematics & Data Sciences, Faculty of Mathematics and Natural Sciences, Andalas University, Kampus Limau Manis, Padang, Indonesia; Department of Political Sciences, Faculty of Social and Political Sciences, Andalas University, Kampus Limau Manis, Padang, Indonesia; Department of Statistics, Universitas Negeri Padang, Jalan Prof. Dr. Hamka, Air Tawar, Padang, Indonesia
