Quality Recognition and Prediction Smarter Pattern Technology with the Mahalanobis-Taguchi System


 「入門MTシステム」の3名の著者がアメリカから出版した書。

 システムの全体像を詳しく解説しており、海外の大学・大学院の

教科書としても採用実績があります。





Recommendation

   We live in uncertain times – uncertain economic times, uncertain political times and times when everyone is worried about what may happen in the future. Local activities can have regional or even global consequences. But the Mahalanobis-Taguchi (MT) system is a beacon of light enabling us to make some sense of this uncertain information and enabling us to make some good predictions.
   The MT system is a diagnostic and predictive method for analyzing patterns in multivariate data that has provided benefits in many diverse applications over the past decade or so. It has proven itself superior in many cases to more traditional artificial intelligence applications such as neural nets.
   Shoichi Teshima, Yoshiko Hasegawa and Kazuo Tatebayashi, a group of eminent Japanese mathematicians and engineers have produced ‘Quality Recognition and Prediction – Smarter Pattern Technology with the Mahalanobis Taguchi System –’ to show how the MT system applies in different ways to benefit a wide variety of industrial, medical and other applications.
  This book discusses the subject in great detail and yet explains it in such a manner that someone not intimately involved with the field can read the book and understand what pattern recognition is and how the MT system works and how it may be applied.
   It starts by giving an explanation of what pattern recognition technology is and how the MT system has been applied to fields as diverse as industrial  applications, medical applications and real estate valuation. It then goes on to discuss the merits of the MT system and how data is handled followed by Unit Space and how it may be applied to weather forecasting and rainfall estimation.
   After discussing how the Taguchi Method has been used for manufacturing yield prediction the book then goes on to show several detailed applications of the Mahalanobis-Taguchi system from industrial applications such as cutting blade wear and machinery monitoring to real estate price prediction.
   Finallythe authors close by showing the differences between the MT system and more traditional approaches such as artificial intelligence using neural nets and statistical methods.
  Yokogawa sees great potential for using the MT system in process automation applications across a wide range of industries for quality management and improved operations.

 Dr. Maurice J. Wilkins
Vice President, Global Strategic Technology Marketing Center, Yokogawa Electric Corporation