A Machine Learning Approach to the Interpretation of Cardiopulmonary Exercise Tests: Development and Validation

Table 3

Summary of the individual probability estimates (%) of the SVM disease classification model.

Conventional clinical diagnosis	Patients’ group	SVM probability estimation (%)
		Mean	SD	Min	Max
CHF	CHF	92.4	10.4	57.0	100.0
	COPD	4.2	6.1	0.0	27.0
	Healthy	3.3	5.7	0.0	19.0
COPD	CHF	10.6	13.2	0.0	47.0
	COPD	79.6	16.9	45.0	100.0
	Healthy	8.7	13.1	0.0	46.0
Healthy	CHF	5.6	7.8	0.0	36.0
	COPD	8.5	11.7	0.0	43.0
	Healthy	85.4	17.1	43.0	100.0

CHF: chronic heart failure; COPD: chronic obstructive pulmonary disease; Healthy: healthy normal participants; SVM: support vector machine; SD: standard deviation; Min: minimum; Max: maximum. Bold numbers denote average probability estimates of the respective group.