Academia.eduAcademia.edu

Figure from:Tracking in Reinforcement Learning

See full PDFdownloadDownload figure
Fig. 1. Boyan Chain: stochastic case

Figure 1 Boyan Chain: stochastic case

Related Figures (3)

Algorithm 1. KTD-V, KTD-SARSA and KTD-Q order approaches, such as residual algorithms [9], the cost function minimized by KTD is thus biased. For the value function evaluation (extension to other cases is straightfor- ward), the bias is:
Algorithm 1. KTD-V, KTD-SARSA and KTD-Q order approaches, such as residual algorithms [9], the cost function minimized by KTD is thus biased. For the value function evaluation (extension to other cases is straightfor- ward), the bias is:
Fig. 2. Boyan Chain: deterministic and non-stationary case
Fig. 2. Boyan Chain: deterministic and non-stationary case
5 Conclusion
5 Conclusion