| 摘要 |
<p>PROBLEM TO BE SOLVED: To easily calculate a ranking transition probability while effectively utilizing the feature of each company by setting a calculation unit according to the attribute information of each company and calculating the probability while using history information in the past when calculating the ranking transition probability by companies. SOLUTION: The ranking transition probability is calculated for calculating the credit risk of a financial institution. Therefore, an attribute to be a reference for the calculation block of ranking transition probability calculation is designated out of attribute information 109 such as class, area, scale and financial information of properties owned by each financial institution by a user, and an average value and the number of pieces of data are calculated concerning attribute information data 110. While using the average value of these attribute information data 110, correlative counting 112 for each attribute is calculated and a calculation block 113 of the ranking transition probability is set from the correlative count and a border value by calculation blocks set by the user. When the number of samples in each calculation block exceeds the prescribed number of samples, a ranking information history obtained from an external ranking organization or internal ranking information is totalized for each calculation block and a ranking transition probability 114 is calculated for each calculation block.</p> |
