Level: Intermediate
【Presenter】
Akira Hososhima, Representative of Tobayashi Consulting
[Presentation summary]
Graph builders and data filters visualize quality data to bring problems into perspective. For example, you can display the number of defective units for each year in a bar graph, and overlay the details of the defect and the supplier on the graph. By graphing the details of defects and the number of units for each year, changes in improvements and deterioration become clear. Based on the graph of year and number of defective units and the graph of defect content and number of units, drop stratification variables into zones such as Group The exclusion function is used to select a specific segment and drill down, but if you exclude by directly manipulating the table, no records will be left, which can lead to misunderstandings later. I do this while saving the script to a data table using a data filter. To check the table of the segment you are interested in, click on the graph to select it, then select Table → Subset to check the details of the table. Graph Builder is also used for model fitting and validation of partition analysis. When checking the fit, use the graph builder to set the main variable or temperature on the X axis, and drop other variables into Group X, etc. This reveals nonlinearities and allows us to discover interactions.
[Presenter profile]
After serving as Director, Director of Research and Development Division, and Director of Quality Assurance Promotion Division at Yamatake Honeywell (currently Azbil), established Tobayashi Consulting. Specialized areas include data analysis, quality improvement, market defect prediction, robust design, experimental planning, design review, human error countermeasures, and on-site guidance. His main publications are ``An Easy-to-Understand Book on Problem Solving'', Nikkan Kogyo Shimbunsha, 2014 (single author). The main paper is ``Proposal for a workshop to promote the dissemination and reception of awareness regarding near misses and discomfort on the production line,'' Quality Control Society 2016 [Quality Technology Award Winner]. Main lectures: "Proposal of a system to visualize organizational factors that cause work errors and promote improvement" (Discovery-Japan 2018) "Solving quality issues using JMP - Case studies of failure analysis and reliability prediction" (Discovery-Japan 2019) ) "Improving the efficiency of experiments and analysis with JMP" (2020) "How to investigate quality issues using text mining" (2021) "Discovering and resolving quality issues with JMP" (2022) "JMP On Air Japan Edition 2021 "Business reform of development, production, and quality assurance using JMP" (1st reliability prediction as basis for handling complaints, 2nd cause investigation using JMP, 3rd problem solving in experimental planning, 4th JMP Management reform/organizational reform, how to proceed with business reform using the 5th JMP)
