Table 6.

Support Vector Machine Classification Results with Demographic Cohorts

	UNIFIED PREDICTION (N=22 PATIENTS)			COHORT (50–64 YEARS) PREDICTION (N=12 PATIENTS)
ACTUAL STATUS	GOLD1	GOLD2	MORE SEVERE	GOLD1	GOLD2	MORE SEVERE
GOLD1	78.54%	19.51%	1.95%	99.24%	0	0.76%
GOLD2	16.96%	65.18%	17.85%	0	87.63%	12.37%
More severe	2.96%	23.47%	73.57%	1.23%	11.6%	87.16%

The unified prediction contains all female test subjects across all age ranges and yields 71.73% accuracy. The cohort prediction focuses on subjects in the 50–64-year age range and yields 89.22% accuracy. Demographics can be used to improve model accuracy. More patients must be analyzed to expand the range of demographics covered by our models, to improve the trained model for predicting health status.

GOLD, Global Initiative for Chronic Obstructive Lung Disease.