BU.450.760 Codebook C1.2 – Credit card defaults

Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

BU.450.760

Codebook C1.2 – Credit card defaults

Credit card defaults data codebook

Source:

Yeh, I. C., & Lien, C. H. (2009). The comparisons of data mining techniques for the predictive    accuracy of probability of default of credit card clients. Expert Systems with Applications, 36(2), 2473-2480.  https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients

Data Set Information:

This dataset was used for research aimed at the case of customers default payments. The key variable in this dataset is “defaultpaymentnextmonth”, a dichotomous variable that indicates whether a customer defaulted on his/her debt. The goal was to generate as accurate as possible out- of-sample predictions for this dichotomic outcome.

Attribute Information:

LIMIT_BAL: Amount of the given credit (NT dollar): it includes both the individual consumer credit and his/her family (supplementary) credit.

SEX: Gender (1 = male; 2 = female).

EDUCATION: Education (1 = graduate school; 2 = university; 3 = high school; and so on … ) MARITAL: Marital status (1 = married; 2 = single; and so on … )

AGE: Age (years).

PAY_t: History of past payment, with t=1 (September 2005), 2 (August 2005), …, 6 (April 2005). The measurement scale for the repayment status is:

•    0 = pay duly

•    1 = payment delay for one month

•    2 = payment delay for two months;

. . .

•    8 = payment delay for eight months;

•    9 = payment delay for nine months and above.

BILL_AMTt: Amount of bill statement (NT dollar), with t=1 (September 2005), 2 (August 2005), …, 6 (April 2005)

PAY_AMTt: Amount of previous payment (NT dollar), with t=1 (September 2005), 2 (August 2005), …, 6 (April 2005)