China
Discovery Summit
Exploring Data | Inspiring Innovation
Beijing | May 18, 2017
Abstracts
The Design of JMP®
John Sall, Co-Founder and Executive Vice President, SAS
JMP was created in response to various forces, situations and opportunities. It evolved to address needs in unique ways. It got very good in certain areas. This is our story of how it all came together over the last 27 years from release 1 to release 12. What were we thinking?
Data Analysis in High-Tech Product Research and Development – Reliability Assessment of Sub-Healthy Products and Systems
Feng-Bin Sun, Senior Staff of Reliability Engineering, Tesla
We are entering an era of big data. Data-driven innovation has enormous economic value, with Big Data product and service sales exceeding $18 billion in 2013 and expected to reach $50 billion by 2017. This presentation will start with an overview of the data analysis trends in many leading innovation companies, especially in product design, development, qualification, production, and maintenance. As a special example of new trends in data analysis, the authors look at the product reliability quantification under sub-healthy conditions – a fractional failure-based reliability assessment methodology. In real-world practice, the sub-healthy condition (fractional failures) can be encountered when:
- Performance degradation has crossed the pre-specified threshold but hasn’t yet manifested as a macro failure (when it ceases to function physically).
- The corrective actions are partially effective (greater than 0 percent, but less than 100 percent).
- The failure analysis cannot duplicate the field failure symptom due to failure diagnosis limitation, etc.
Examples are given using JMP software to illustrate data collection, failure classification and fractional failure determination, data entry format, life distribution parameter estimation, reliability quantification, and field risk prediction. It is believed that this talk will be beneficial to a wide audience, including reliability practitioners, theorists and management.
Data Science is Sexy, but Numbersense is Priceless
Kaiser Fung, Author, Numbersense and Numbers Rule Your World
Analyzing data is like running an obstacle course. The path is laid with trapdoors, dead ends and diversions. The best analysts count on a keen sense of direction as they navigate data. This numbersense is the priceless asset in data science. I will describe some recent analyses that went off track, and discuss how to prevent falling into such traps. Troubleshooting using statistical know-how is the easy part; sensing fault lines and piecing together root causes prove to be much more challenging. Examples will be drawn from diverse fields, such as predictive analytics, A/B testing, advertising technology, Web scraping and Open Data projects.
-
Applying Statistical Modeling to Crop Research
Chongfa Yang, Professor, Hai Nan University
Crop research is complex and requires advanced statistical methods. With the help of modern analysis software like JMP, crop research will be easier. Using examples from different kinds of crop case studies, this presentation will outline a comprehensive introduction to statistical modeling, concrete analysis methods and skills, and the end results.
-
Drive Continuous Improvement in Data Analysis
Qiang Wu, Lean Six Sigma Internal Consultant, Shenzhen China Star Optoelectronics Technology Co., Ltd.
Shenzhen China Star Optoelectronics Technology (CSOT) ramped up production and introduced Six Sigma and JMP in 2012. As the era of rapid changes and product iteration in the market is approaching, production efficiency and quality have become a key issue. Nothing great comes to be all at once, the same goes for high efficiency and high quality improvement. The continuous improvement of both attitude and ability is essential. With the help of JMP, CSOT evolved from "relying on experience" to "relying on data." Both engineers and managers no longer express words like "I think," and are more likely to use words like "data" and "variation." Data analysis has been integrated into the corporate culture. During the improvement of process, we use the Six Sigma method to complete all kinds of analyses and to solve a large number of complex problems. Data mining for further information will be our next focus.
-
JMP® Provides Assistance for the Smooth Mass Production of the Nation's First IC Package Substrate
Xingyong Zhang, Six Sigma Master Black Belt, AT&S
Austria Technologie & Systemtechnik AG (AT&S) is a global leader in high-end printed circuit board manufacturing. AT&S (Chongqing) Co. Ltd. will mass produce 7um - 15um width semiconductor package substrates in the future, specifically to support the 10nm - 40nm wafers and 25um - 50um printed circuit boards. There is a high barrier to entry for semiconductor package substrate. The AT&S plant located in the Liangjiang New Area of Chongqing is one of only three new generation advanced semiconductor package substrate manufacturers in the world, and the only one in China. As the first domestic program that introduced semiconductor package substrates， the entire project has over 100 control parameters and nearly 100 defect opportunities. JMP data visualization and analysis has become an indispensable segment of the project. We will highlight and introduce how JMP software data visualization and the personalized production site information display platform can boost the smooth production of the country's first IC package substrate.
-
Multiple Correspondence Analysis – A New Platform for Categorical Variables, Ready to Be Explored!
Jianfeng Ding, JMP Senior Research Statistician Developer, SAS
In multivariate analysis, dimension reduction into a small number of factors is the most important step for capturing the variability among a large number of variables. In JMP®, we have the Principal Components (PC) platform to do dimension reduction for continuous variables; however, when we have categorical variables, we cannot model using the PC platform. In JMP 12, we added the Multiple Correspondence Analysis (MCA) platform, which takes multiple categorical variables as input variables and seeks to identify associations between levels of those variables. MCA, a data analysis technique popular in Europe and Japan, is now an addition to our already-robust multivariate toolbox. In this presentation we will use the Le Roux’s Taste data set combined with survey data collected from our JMP division employees to explore data preparation, statistical analysis, graphical representation and interpretation using the MCA platform. We will also discuss various topics in MCA, such as cloud of categories, cloud of individuals, distances, dimensionality, contributions and supplementary elements to help support the analysis.
-
Process Optimization in High-Tech Industry With DOE
Frank Yang,Director of Technical Experts,SMIC
Design of experiments （DOE) is a very powerful tool. However, many researchers and technology managers are ignorant of DOE or only apply it mechanically. They do not know that in addition to the classic DOE methods, there are more effective, flexible, and advanced design methods and techniques. This lecture starts with a virtual experiment to introduce the core concepts of DOE, and then gradually goes deeper to perform in-depth analyses, as well as clarify common erroneous understandings while sharing efficient learning methods to help everyone get rid of the various obstacles of in-depth DOE learning.
-
Using JMP® to Achieve the Organic Combination of Virtual Product Technology and Big Data Analysis
Lawrence Wu, Manager, R&D, FemCare Modeling & Simulation, Procter & Gamble
Whisper is a feminine health product brand of the Procter & Gamble Company in the United States. From its first day in China, Procter & Gamble's Whisper brand has strived to provide advanced, high-quality, and comfortable personal hygiene care products for women in China. The tightness of the packaging for sanitary pads is a key indicator to customer focus attraction. Packaging tightness is affected by the early end of the production process, and the early end of the production process cannot be easily modified. For this particular situation, Procter & Gamble has established the groundbreaking virtual product packaging design concept. The company combined the finite element model, the experimental design/fitting of JMP software, and the Monte Carlo simulation function to perform virtual packaging R&D. The process can also predict and improve factors that affect product quality and significantly enhance the quality of packaging for the pad category products.