Skip to the content.

Data-Driven Model Reduction, Scientific Frontiers, and Applications

Doowon Choi, Department of Industrial & Systems Engineering
Correlation Trees with Application to Neural Correlates Studies

Abstract

Brain-behaviour relationship is the focus in neural correlates studies. The relationship is characterized by correlation between brain activation responses (e.g., change in brain blood flow) and human behaviour measures. Such correlation depends on some subject-related covariates such as age and gender, so it is necessary to identify subgroups within the population that have different brain-behaviour correlations. However, the subgrouping is made by manual specification in current practice, which is inefficient and may ignore potential covariates whose effects are unknown in the literature. Following the spirit of decision trees, we propose an approach called correlation tree to identify subgroups in an automatic, optimal manner. In constructing a correlation tree, the split variable at each node is selected through an unbiased variable selection method based on partial correlation test, and the optimal cutpoint of the selected split variable is determined through exhaustive search under a pre-defined objective function. The proposed approach is applied to a real dataset on risk-decision making.