抄録
The objective of this study is to improve overall judgment of a health examination using Mahalanobis distance by summarizing the correlation of various types of data, instead of observing the data item by item as is currently done. In this study, only the data such as gender, age, the questionnaires of previous and this year, urine tests and blood tests of only previous year were used and predicted the health condition of this year. For the reliability improvement of final diagnosis, an orthogonal array was used to assign the items which were used in the beginning of the study. The effects of these items on Mahalanobis distance were observed and insignificant effects were screened out to improve the efficiency of final diagnosis.
