| Summary: | 碩士 === 明新科技大學 === 工業工程與管理研究所 === 97 === Abstract
Process Capability Indices (PCIs) are important tools for the manufacturing industry to examine whether the manufacturing process meets the product specifications. The main function of PCIs is to effectively control the manufacturing process to produce high-quality products. To check whether the manufacturing process of a certain company meets the intended standards, this research used SAS IML program to implement Monte Carlo, Parameter Simulation, and Bootstrap re-sampling methods, and to test the hypotheses versus . The main purpose was to examine the practicality of using sampling distributions of obtained by parameter simulation and bootstrap methods to test the above hypotheses.
The distributions considered were Normal, Uniform, and Gamma. The results showed that：(1)parameter simulation performed better than bootstrap did, (2)the correct test proportions of bootstrap on normal, uniform, or Gamma had no significant differences, (3)the correct test proportions of parameter simulation on Gamma were higher than those on normal and Uniform, and (4) parameter simulation and bootstrap performed worse with small samples.
Keywords：Process Capability Indices, Parametric Simulation, Bootstrap.
|
|---|
