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

Brain Informatics

Table 8 Feature significance with respect to Health-score using the permutation importance approach

Health-score	Regression	KNN	SVM	Decision tree	Random forest	Gradient boost	Average
Gender	0.026	0.059	0.030	0.043	0.046	0.039	0.040
Ethnics	0.051	0.062	0.040	0.063	0.058	0.049	0.054
Education	0.005	0.062	0.015	0.057	0.046	0.019	0.034
Relationship	0.037	0.063	0.033	0.062	0.056	0.066	0.053
Adversity	0.068	0.042	0.070	0.043	0.040	0.056	0.053
Exercise	0.170	0.094	0.159	0.090	0.130	0.167	0.135
Alcohol	0.006	0.046	0.021	0.043	0.036	0.020	0.029
Tobacco	0.046	0.042	0.030	0.043	0.034	0.040	0.039
Relation-Impact	0.081	0.064	0.135	0.065	0.066	0.073	0.081
Communication	0.092	0.075	0.095	0.069	0.071	0.069	0.078
Therapy	0.153	0.074	0.114	0.095	0.095	0.132	0.111
Medication	0.043	0.063	0.076	0.060	0.054	0.048	0.057
Health-service	0.077	0.066	0.095	0.068	0.067	0.059	0.072
Social-distancing	0.028	0.031	0.013	0.042	0.032	0.044	0.032
Risk-group	0.049	0.050	0.014	0.048	0.061	0.050	0.045
Living-group	0.025	0.060	0.042	0.068	0.065	0.033	0.049
Contract-risk	0.042	0.047	0.016	0.041	0.041	0.036	0.037