Rebalancing Docked Bicycle Sharing System with Approximate Dynamic Programming and Reinforcement Learning

Table 4

Mean and standard deviation by prioritizing strategies.

Strategy	50%		100%		150%
	Mean	Standard deviation	Mean	Standard deviation	Mean	Standard deviation
Strategy 1: all stations	92.95	1.74	93.54	4.11	94.16	2.66
Strategy 2: near stations	94.06	1.08	92.27	1.08	91.76	3.5
Strategy 3: stations with large errors	98.56	1.73	93.70	1.97	90.87	1.45
Constraint-free strategy	96.46	2.19	96.93	4.99	96.85	3.84