Incremental gradient descent imputation method for missing data in learning classifier systems

Learning with incomplete or missing data has been a major challenge in learning classifier system. One method for covering missing data is imputing missing values based on the statistic of known values. Another is marking them matching arbitrary case. A new approach using incremental gradient descent imputation model is proposed in this paper, which use the relationship among variables to estimate the missing value. And, some experiments are conducted in order to compare the performance of new approach and other classical covering methods.