EPJ Data Science

Table 4 A relative number of cases when prediction accuracy of four different prediction algorithms surpasses the predictability limit calculated for the two types of mobility sequences of various spatio-temporal resolution

From: Explaining human mobility predictions through a pattern matching algorithm

	Next-place [%]			Next time-bin [%]
	1688 m	204m	33 m	1688 m	204m	33 m	1-hour	30-minutes
GRU	25.67	5.33	3.33	26.50	10.33	8.00	16.11	13.78
RF	25.30	6.33	6.67	27.00	11.11	8.50	17.44	13.67
MC	18.00	6.00	5.33	18.17	3.33	3.50	8.44	16.00
Toploc	18.00	0.67	4.33	21.83	8.33	8.33	8.22	10.00

GRU is the deep neural network, RF is an ensemble decision trees method, MC is the Markov chains-based prediction algorithm, and Toploc is a baseline predictor. Bold values indicate the highest result for each sequence type.

Back to article page