CreditData: About 20,000 fictional credit records.

In xxzcool/scoremodel: Credit Scoring Modeling Toolbox

Description

A dataset containing the personal and credit attributes of almost 20,000 fictional customers. The variables are as follows:

Usage

data(CreditData)

Format

A data frame with 19805 rows and 23 variables

Details

• gender. gender of customer (F, M)

• marriage. marriage of customer (unmarried, married, divorced)

• education. education of customer (primary, middle, high, polytechnic, junior, bachelor)

• housing. housing of customer (unit, rent, parents, own)

• loanamount. loan amount (500–10800)

• loanperiod. period of loan (3–24)

• queryid. suspect query reuslt by id number (positive, negative)

• queryphone. suspect query reuslt by phone number (positive, negative)

• queryhis15d. suspect query history in the last 15 days (positive, negative)

• queryhis30d. suspect query history in the last 30 days (positive, negative)

• queryhis90d. suspect query history in the last 90 days (positive, negative)

• applyflag. whether applied before (0, 1)

• multiloantimes. times of multiple loan in the last 12 months (1–31)

• purchasingPI. purchasing power index of customer (10–100)

• rationalCI. rational consumption index of customer (0–100)

• equipmentBI. equipment behavior index of customer (302–848)

• transactionRI. transaction risk index of customer (359–818)

• consumptionCI. consumption composite index of customer (350–850)

• gscore. customer's score graded by credit institution G (443–703)

• kscore. customer's score graded by credit institution K (300–850)

• bscore. customer's score graded by credit institution B (365–854)

• pscore. customer's score graded by credit institution P (304–789)

• target. response variable, 1 means bad customer, 0 means good customer (0, 1)

xxzcool/scoremodel documentation built on May 4, 2019, 10:56 a.m.