Skip to main content

Advertisement

Table 4 Feature importance for the fitted augmented Logistic Regression model for week 20

From: Applying machine learning to predict future adherence to physical activity programs

Feature Coefficient P-value
Intercept 1.378 0.033
Week number −0.122 < 0.001
Initial average daily steps −0.001 < 0.001
Average daily steps 0.0008 < 0.001
Last week average daily steps 0.0004 < 0.001
Initial MVPR morning −0.029 0.138
Initial MVPR afternoon 0.027 0.041
Initial MVPR evening 0.015 0.242
Average MVPR morning 0.067 0.02
Average MVPR afternoon −0.037 0.103
Average MVPR evening 0.005 0.805
Last week MVPR morning −0.032 0.058
Last week MVPR afternoon 0.001 0.920
Last week MVPR evening −0.003 0.797
Initial MVPA intensity −0.158 0.479
Average MVPA intensity −0.354 0.280
Last week MVPA intensity 0.515 < 0.001
Average goal achieving percentage 0.161 0.849
Last week goal achieving percentage 0.294 0.340