Penyelesaian masalah manajerial dengan metode iterasi kebijakan pada discounted markov decision prosesses

Abstract

The aim of this paper is to solve a amanagement problem in advertising and production that have been formulated into a stochastic model of the decision-making system. The Decision-making technique used in this paper is the Markov decision process with policy iteration method. To determime the optimal policy, it employed the discounted-return policy improvement algorithm that uses a discount factor equal to 0.9 for each case. In the case of advertising, the manufacturers will use RCTI as a promotion media when their product sales are low and Indosiar when they have a good sales. When they have satisfied sales, they use SCTV as media promotion. In the case of production management, farmers must keep the fishes in the pond at least 2 tons when they harvest their fishes.