Seven decision tree algorithms for missing values were compared based on split point selection, split variable selection, and the predictive accuracy. The simulation experiments were carried out under two scenarios. We found that the performance of missing value algorithms depends on the characteristic of data set under the analysis and none can outperform others in all cases.
|Title of host publication||Statistical Data Mining and Knowledge Discovery|
|Number of pages||18|
|Publication status||Published - 2003 Jan 1|
All Science Journal Classification (ASJC) codes
- Economics, Econometrics and Finance(all)
- Business, Management and Accounting(all)
- Computer Science(all)