Key points are not available for this paper at this time.
Control of an output value y when it is a function of various variables such as A, B, is usually achieved by controlling those causes A, B. After discussing methods using partial derivatives and Monte-Carlo, which are either inaccurate or inefficient, the use of orthogonal arrays, and the use of analysis of variance for the data calculated or conducted by the arrays, are proposed in this paper as these methods have no such defects. A live example taken from an electric circuit used for a colour T.V. set is used for the explanation.
Genichi Taguchi (Wed,) studied this question.