Estimating Method for Variation of Function of Representing Value for Probability Variable, Computer Program and Recording Medium Recorded the Program
To allow estimating variations of functions of representing values for probability variables within a time shorter than a conventional batching method even if a data set for use is relative small in size. SOLUTION: In an estimating method for variation of function of representing value for probability variable, an average value E [X] is decided as a representing value for probability variables X and also statistical amounts of the variables X (e.g. a standard deviation and a confidence interval) are decided. Moreover, a gradient f’ of the decided average value E [X] of a performance function f (a function linking the variables X with a performance value Z) as a function of the variables X is acquired to convert the acquired statistical amounts into the function f by using the acquired gradient
- File Name: Roser-2002-Estimating-Method-for-Variation-of-Function-of-Rep.pdf
- File type: pdf
- File size: 1.08 MB
- Date: May 11, 2014
- SEO Title: Estimating Method for Variation of Function of Representing Value for Probability Variable, Computer Program and Recording Medium Recorded the Program
- SEO Description: To allow estimating variations of functions of representing values for probability variables within a time shorter than a conventional batching method even if a data set for use is relative small in size. SOLUTION: In an estimating method for variation of function of representing value for probability variable, an average value E [X] is decided as a representing value for probability variables X and also statistical amounts of the variables X (e.g. a standard deviation and a confidence interval) are decided. Moreover, a gradient f' of the decided average value E [X] of a performance function f (a function linking the variables X with a performance value Z) as a function of the variables X is acquired to convert the acquired statistical amounts into the function f by using the acquired gradient
- SEO Keywords: