Evaluation of different radon guideline values based on characterization of ecological risk and visualization of lung cancer mortality trends in British Columbia, Canada

Table 1 The classification metrics for each balanced random forest algorithm. Accuracy is defined as the proportion of an observed class that was correctly classified. Precision is defined as the proportion of a predicted class that was correctly classified. Kappa can be interpreted as the percent improvement in overall accuracy of a classifier compared with the expected overall accuracy of a random classifier

	Threshold in Bq m⁻³	Lower-than-threshold Accuracy	Lower-than-threshold Precision	Higher-than-threshold Accuracy	Higher-than-threshold Precision	Kappa	Kappa Gain
a)	600	0.81	0.97	0.69	0.22	0.25	0
b)	500	0.83	0.96	0.72	0.32	0.36	0.11
c)	400	0.83	0.96	0.74	0.37	0.39	0.03
d)	300	0.8	0.94	0.73	0.42	0.41	0.02
e)	200	0.8	0.91	0.76	0.55	0.49	0.08
f)	150	0.77	0.88	0.76	0.6	0.5	0.01
g)	100	0.79	0.86	0.83	0.75	0.61	0.11
h)	50	0.77	0.8	0.86	0.84	0.63	0.02

ISSN: 1471-2458