Apply Big Data Analysis Technique with the Semiconductor Device Plating Machine Maintenance and Repair Data for Yield Improvement

碩士 === 國立高雄應用科技大學 === 電機工程系博碩士班 === 104 === In the history of the semiconductor industry in recent years, in 2008 IBM announced the copper process developed, IC process technology from aluminum process technology to copper process technology, announcing the coming copper Cheng era, so equipment man...

Full description

Bibliographic Details
Main Authors: YEH,PEI-CHIANG, 葉沛強
Other Authors: Chung-Hong Lee
Format: Others
Language:zh-TW
Published: 2016
Online Access:http://ndltd.ncl.edu.tw/handle/88591648159975625762
Description
Summary:碩士 === 國立高雄應用科技大學 === 電機工程系博碩士班 === 104 === In the history of the semiconductor industry in recent years, in 2008 IBM announced the copper process developed, IC process technology from aluminum process technology to copper process technology, announcing the coming copper Cheng era, so equipment manufacturers but also because of the development of this technology without segments new device to correspond to the new era, the present study to establish a personnel equipment plating equipment maintenance and maintenance analysis model, combined with changing factor data, as the plating equipment maintenance and maintenance of early warning systems, is the main purpose of this research work. In this study, we collected data related to machine maintenance and troubleshooting from 2014 to 2016, and used the relevant data as the criteria for classification. In the choice of variables, through correlation analysis, confirm the homogeneity of variables , This paper will focus on the use of multi-level support vector machine (SVM) model, in order to illustrate the contents of my research in the information and the feasibility of the proposed method. And with the difference between the proposed relationship between the plant to illustrate the data can do under the pre-failure and failure to do the case of rapid intervention and reminders in the data correlation analysis to find the impact of quality variables to variables Explain the conceptual approach and effectiveness of large data analysis.