Skip to main content

Table 6 Feature significance with respect to BRS6 using the permutation importance approach

From: Investigating the mental health of university students during the COVID-19 pandemic in a UK university: a machine learning approach using feature permutation importance

BRS6

Regression

KNN

SVM

Decision tree

Random forest

Gradient boost

Average

Gender

0.060

0.067

0.064

0.069

0.090

0.080

0.072

Ethnics

0.184

0.074

0.079

0.089

0.078

0.116

0.103

Education

0.055

0.068

0.060

0.055

0.061

0.052

0.058

Relationship

0.014

0.077

0.080

0.064

0.085

0.057

0.063

Adversity

0.087

0.053

0.065

0.066

0.070

0.079

0.070

Exercise

0.052

0.065

0.077

0.071

0.062

0.051

0.063

Alcohol

0.015

0.044

0.036

0.036

0.038

0.034

0.034

Tobacco

0.017

0.039

0.041

0.037

0.032

0.043

0.035

Relation-Impact

0.015

0.066

0.070

0.058

0.059

0.043

0.052

Communication

0.108

0.070

0.078

0.079

0.082

0.096

0.086

Therapy

0.103

0.061

0.058

0.055

0.054

0.045

0.063

Medication

0.026

0.055

0.057

0.055

0.044

0.046

0.047

Health-service

0.124

0.065

0.057

0.069

0.070

0.078

0.077

Social-distancing

0.023

0.029

0.031

0.025

0.023

0.030

0.027

Risk-group

0.042

0.053

0.045

0.063

0.050

0.044

0.050

Living-group

0.018

0.064

0.058

0.064

0.056

0.068

0.055

Contract-risk

0.057

0.051

0.044

0.045

0.046

0.039

0.047

  1. The top-5 most significant features highlighted in bold