Penyelesaian Model Tahap Terhingga dan Takhingga pada Proses Keputusan Markov dan Aplikasinya di Bidang Pertanian.

Abstract

Markov decision process is a decision making process using Markov chain for stochastic models. The aim of this paper is to formulate a stochastic model involving states, actions and rewards. Further, the model is applied into agricultural sector, especially on determination of the optimal revenue based on actions specified. Also to determine an optimal policy that maximizes the reward. The methods used in this study are the complete enumeration, the policy iteration and the linear programming methods. Among the methods used, the most efficient method is the policy iteration. Based on the data used, determination of the policy using those three methods concluded that farmers would not use fertilizer when the soil fertility is good, and will use fertilizer when the soil fertility are moderate or low. Especially, for the case of the discount 0.7, the agricultural problem resulting the same policy with the case of no discount rate.