PROBLEM TO BE SOLVED: To accurately predict assembly yields of a product.;SOLUTION: An assembly yield prediction device 1 has: an acquisition part; a multivariate distribution calculation part 211; and a prediction part 22. The acquisition part acquires multivariate distribution parameters 56, 63 of partial dimensions including correlation information between the partial dimensions in the respective components. The multivariate distribution calculation part 211 calculates multivariate distributions 57, 67 of check part dimensions based on the multivariate distribution parameters 56, 63 of the partial dimensions and sensitivity information 53. Here, the sensitivity information 53 shows a degree of influence to be given to the dimensions of a plurality of check parts in a product by variation of the respective partial dimensions. In addition the multivariate distribution parameters 57, 67 of the check part dimensions include the correlation information to be generated between the dimensions of the plurality of check parts resulting from a fact that the variation of the respective partial dimensions in the same component gives the influence to the dimensions of the plurality of check parts. Then, the prediction part 22 predicts assembly yields 60, 68 of the product based on the calculated multivariate distribution parameters 57, 67 of the check part dimensions.;COPYRIGHT: (C)2015,JPO&INPIT
展开▼