IMPLEMENTASI MARKOV CHAIN MONTE CARLO PADA PENDUGAAN HYPERPARAMETER REGRESI PROSES GAUSSIAN

Abstract

This paper studies the implementation of Markov Chain Monte Carlo on estimating the hyperparameter of Gaussian process. Metropolish-Hasting (MH) algorithm is used to generate the random samples from the posterior distribution that can not be generated by a direct simulation method. This algorithm require only a proposal distribution for generating a candidate point. In this paper uniform distribution is choosen as the proposal distribution.