eliasmei / dlm

KDD20-DLM

Code release of "Imputing Various Incomplete Attributes vid Distance Likelihood Maximization" (KDD 20).

Parameters

The input and output of DLM algorithm is:

Method

setParams(K, L, Can);

Input:

int K;  // the number of considered largest likelihoods 
int L;  // the number of learning neighbors for distance models
int Can;  // the number of candidates for imputation

Output

HashMap<Position, Cell> cellMap

Library

jama.jar is used to implement the code.

Citation

If you use this code for your research, please consider citing:

@inproceedings{DBLP:conf/kdd/SongS20,
  author    = {Shaoxu Song and
               Yu Sun},
  title     = {Imputing Various Incomplete Attributes via Distance Likelihood Maximization},
  booktitle = {{SIGKDD}},
  pages     = {535--545},
  publisher = {{ACM}},
  year      = {2020}
}