dtgini: Runs a Decision Tree Gini Index and Decision Tree Information...

Description Usage Arguments Value

View source: R/dtgini.R

Description

This procedure uses each feature to perform a one-node split of all N samples in the parent node into 2 daughter nodes using N-1 cut points, as in a decision tree. This procedure's goal is to find the most optimal split to identify between cases and controls for each feature. To do this, this function identifies the smallest weighted Gini Index value and smallest weighted Information Gain value for each feature.

Usage

1
dtgini(dat)

Arguments

dat

a list containing 3 elements: case, a list of case/control statuses; feat, a matrix of normalized feature data; maxfeat, a list of max features from each column in feat

Value

finalresults a list of Gini Indices


abcsFrederick/BMDK documentation built on Aug. 9, 2021, 2:19 p.m.