Credit Risk Prediction Using Fuzzy Immune Learning

Table 2

Parameter specification of IFAIS in our experiments.

Parameter	Value in Australian	Value in German
initial Population Size	100	300
max Iteration	50	50
default Age	5	5
selection Size	100	100
clone Number	10	10
max Term Changes Number	3	2
dontCare Replacement Rate	0.5	0.2
max Rule Set Size	5	10
accuracy Threshold	0.03	0.03
max Rule Similarity Length	10	17
min Covered Percent	0	30
memory Weight	0.5	0.5
𝑤𝑃	0.01	0.35
𝑤NCP	0.69	0.1
𝑤𝑁	0.01	0.45
𝑤NMP	0.29	0.1
𝑤BF	0.8	0.999
𝑤LEN	0.2	0.001