PRICE AND LEAD TIME DECISION MODEL DEVELOPMENT USING PARTICLE SWARM OPTIMIZATION ALGORITHM IN MAKE TO ORDER SYSTEM FOR FASHION INDUSTRY

• Main Author: Zahabiyah, Rifdah
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/63882
• Institution: Institut Teknologi Bandung

This research focuses on selling fashion products using the MTO production system or known as the Pre-Order (PO) sales strategy. Often, PO sales were chosen to avoid deadstock phenomenon due to rapid changes of tren without considering the suitability of the product with the chosen strategy, thus impacting in low demand. One of the research objectives is to develop a prediction model of PO/non-PO sales strategy based on 12 input variabels that explain the product characteristics. The method used is a hybrid Artificial Neural Network (ANN) and logistic regression, called ANN-Plogit. Logistic regression used to determine the significance of each predictor to the target and the ANN model used as predictions. Evaluation of model performance is based on error rate (MSE) and confusion matrix. The result is the model of ANN-Plogit 6 variabels has the best performance with the lowest prediction error and the highest prediction accuracy. In the application of MTO, the amount of lead time is a crucial decision which is often compensated by a lower price in order to obtain an increase in the number of demands. This research also discusses the pricing and lead time decision model in the MTO system by considering the policy of minimum order quantity (MOQ) that set by the production vendors. This condition allows for inventory costs even for MTO production. The decision variabels are price and lead time as well as the number of demands that optimized simultaneously to maximize sales profit. The solution search is carried out using a metaheuristic approach, the Particle Swarm Optimization algorithm is chosen because of its simplicity and the computation time tends to be short. Based on a search with the PSO Algorithm, the price of 324,750 and a lead time of 9 days considered as the best solution with a maximum profit of IDR 80,526,000. Through the validation of the results, this decision is considered feasible to be applied as pre-order sales in terms of market preferences.