Table 1. Performance measures (AC, accuracy; FS, F1 score; KP, Kappa statistic) across all classifiers and class distributions for dynamic data, forward roadway vs. center stack.

	Original dataset			Balanced dataset
	AC	FS	KP	AC	FS	KP
k-nearest neighbor	0.93	0.19	0.16	0.80	0.80	0.59
Random forest	0.86	0.30	0.24	0.79	0.78	0.59
Multilayer perceptron	0.78	0.29	0.22	0.80	0.82	0.59
Hidden Markov model	0.84	0.28	0.22	0.83	0.68	0.57