Author disclosed the reports to robust design on optimizing tool which was showed at the 3rd Pacific Rim Statistical Conference for production Engineering Dec 13/14 2018 at Taipei Taiwan: National Tsing Hua University. It was originated on 1980’s at Japan as Taguchi methods.

  It was mainly expanded to engineers by researching through the case studied in many engineering fields’ likes’ chemicals, electronics, casting, copy, finishing etc. At near 2000’s, there were known to have been the estimation problems for optimum confirmation trials. By statistic survey, there were 62% cases which optimum conditions could not be exceeded the max value of the original data-set with the normal analysis procedure. Engineers had started to realize there were something wrongs in the process as methodology which were proposed by Taguchi. Taguchi asked the linear relation for analysis; however, engineers recognized all of the researching subjects have non-linear effects. Also, SN-ratio which was supplied by Taguchi has been created non-linear effects at transforming raw data to as design indicator. They will contaminate the main effects; finally, engineers will fail in selecting the real optimum conditions.

   Especially, Taguchi had been sending the best-case studies to outside oversea including USA after screening for collecting cases. So, it became to be late that statisticians realize there were many mathematic conflictions to miss the optimum conditions. It started to open/show the own raw case studies to discuss/exchange the optimum-information’s between engineers from the end of 1990’s.

   Statisticians criticized that at the tuning process stage to target will not be maintained variation as PerMIA (Performance Measures Independent of Adjustment). Also, we confirmed there were different the optimum conditions based upon between the compound and orthogonal array noise factor.

   It is the time to discuss with the inspection report for the SN-ratio optimum prediction accuracy of Taguchi two step design.

Keywords: Robust Design; Taguchi Methods; SN Ratio; Two Step Design; Tuning; PerMIA; Optimum; Statistics

References

1. Taguchi G and Wu Y. Introduction to off-line Quality Control, Japan Quality Control association (1980).
2. Teruo MORI. “The 62% Problems of SN Ratio and New Conference Matrix for Optimization”. Proceeding of the pacific Rim Statistical Conference in Korea for Production Engineering (2018): 15-20.
3. Teruo Mori. “Survey for Optimum the difference prediction and actual value” (Taguchi is OK??) No113 Proceeding 16th Quality Engineering Association (2008).
4. Jeff: Sober Ghosh. Statistical Design and Analysis of industrial Experiment Marcel Dekker, NC (USA) (1990).
5. Teruo MORI., et al. “Comparison of the behaviors and difference of the optimum between compound and orthogonal arrays”. 48.1 (2018): 105-111.
6. Teruo MORI. “Survey and analysis for Reverse data to compound noise factors”. No2-1 117th JSQC Nagoya (2018). 
7. Leon., et al. “Performance Independent of Adjustment: An explanation and Extension of Taguchi’s Signal-to-Noise Ratio”. Technimetrics (1987): 253-285.
8. X Shirley Hou. “On the use of compound noise factor in parameter design experiment” Applied Stochastic Models in Business and Industry 18 (2002): 225-243.
9. Shun Matsuura. “Precision of the Estimation of the Response Mean and the Response variance in Compound Noise Experiments”. Japanese society for Quality control 44 (2014): 123-134.
10. Data base (Excel) for 62% problems.

Citation

Citation: Teruo Mori. “The Inspection Report for the SN-Ratio Optimum Prediction Accuracy of Taguchi Two Step Design. It is Grade D?" Acta Scientific Nutritional Health 4.4 (2020): 88-98.

Copyright

Copyright: © 2020 Teruo Mori. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.