This study proposes a two-phase research framework to address the problem of capacity-demand mismatch in the high-tech industry. The first phase builds demand forecast models such as linear regression and autoregression models. It also models and employs the latent information (LI) function for generating the virtual data which benefits the data learning process of the neural network for predication enhancement. Based on the resulting forecast demand scenarios, the second phase focuses on the capacity decision and investigates the regrets of capacity surplus and capacity shortage. We compare the expected value (EV) solution, the minimax regret (MMR) approach, and the stochastic programming (SP) technique which support the capacity decision. We conduct an empirical study of a TFT-LCD firm to validate the proposed framework. From the results, we conclude that the proposed framework, in particular the SP technique, provides a robust capacity level addressing the problem of capacity-demand mismatch.
All Science Journal Classification (ASJC) codes
- Computer Science(all)