Context Transfer in Reinforcement Learning Using Action-Value Functions

Figure 6

The comparison of average reward of learning with and without transfer for crossroad traffic controller.