The Effects of Missing Data Imputation and Discretization Techniques on Learned Bayesian Networks
October 19, 2016
Track: ACM SRC
In this work, we want to determine whether Bayesian networks learned from imputed and discretized data differ from ground truth Bayesian networks, and whether this difference varies depending on the imputation and discretization methods used. Our main contribution is an early study on determining whether different imputation and discretization methods affect learned Bayesian networks.