Trinary Decision Trees for handling missing data

TL;DR

This paper presents the Trinary decision tree, a new algorithm for handling missing data in decision trees that performs well in MCAR scenarios and introduces a hybrid model for broader robustness.

Contribution

The paper introduces the Trinary decision tree and a hybrid TrinaryMIA model, advancing missing data handling without assuming missingness contains information.

Findings

Outperforms in MCAR missing data scenarios

Lags behind in Informative Missingness (IM) scenarios

Hybrid TrinaryMIA model is robust across missing data types

Abstract

This paper introduces the Trinary decision tree, an algorithm designed to improve the handling of missing data in decision tree regressors and classifiers. Unlike other approaches, the Trinary decision tree does not assume that missing values contain any information about the response. Both theoretical calculations on estimator bias and numerical illustrations using real data sets are presented to compare its performance with established algorithms in different missing data scenarios (Missing Completely at Random (MCAR), and Informative Missingness (IM)). Notably, the Trinary tree outperforms its peers in MCAR settings, especially when data is only missing out-of-sample, while lacking behind in IM settings. A hybrid model, the TrinaryMIA tree, which combines the Trinary tree and the Missing In Attributes (MIA) approach, shows robust performance in all types of missingness. Despite the potential drawback of slower training speed, the Trinary tree offers a promising and more accurate method of handling missing data in decision tree algorithms.