Skip to main content

Table 2 Model names and specifications of the six ensemble models

From: Predicting later categories of upper limb activity from earlier clinical assessments following stroke: an exploratory analysis

Model

# Trees

Predictors Included in Input Data Set

# Predictors considered at each node (m =)*

UL clinical measures

Non-motor clinical measures

Demographics

Small bagged

2000

✔

✗

✗

3

Small random forest

2000

✔

✗

✗

√3 = 2

Medium bagged

2000

✔

✔

✗

7

Medium random forest

2000

✔

✔

✗

√7 = 3

Large bagged

2000

✔

✔

✔

17

Large random forest

2000

✔

✔

✔

√17 = 4

  1. *m is a tuning parameter for the bagged trees and random forest models