| URL: | https://github.com/zislam/DMI/ | |
| Author: | Michael Furner | |
| Maintainer: | Michael Furner <mfurner{[at]}csu.edu.au> |
Class that implements the DMI imputation algorithm for imputing missing values in a dataset. DMI splits the dataset into horizontal segments using a C4.5 (J48) decision tree in order to increase the correlation between attributes for EMI. EMI is performed to impute missing numerical attribute values and mean/mode (within a leaf) imputation is used to perform missing categorical attribute values. Uses Amri Napolitano's EMI implementation for Weka.