Applying Clustered KNN Algorithm for Short-Term Travel Speed Prediction and Reduced Speed Detection on Urban Arterial Road Work Zones

Table 4

Mean absolute percentage error (MAPE), mean absolute error (MAE), and root mean square error (RMSE) according to K values when lag duration is 20 minutes.

K	MAPE (%)	MAE	RMSE
1	7.54	2.35	2.66
2	7.02	2.20	2.48
3	7.02	2.21	2.49
4	6.97	2.23	2.50
5	6.90	2.22	2.48
6	6.97	2.26	2.52
7	6.98	2.28	2.53
8	7.05	2.29	2.54
9	7.12	2.32	2.57
10	7.26	2.37	2.61