Measuring Structural Similarities in Finite MDPs

  title={Measuring Structural Similarities in Finite MDPs},
  author={H. Wang and Shaokang Dong and Ling Shao},
  • H. Wang, Shaokang Dong, Ling Shao
  • Published in IJCAI 2019
  • Computer Science
  • In this paper, we investigate the structural similarities within a finite Markov decision process (MDP). We view a finite MDP as a heterogeneous directed bipartite graph and propose novel measures for the state and action similarities, in a mutually reinforced manner. We prove that the state similarity is a metric and the action similarity is a pseudometric. We also establish the connection between the proposed similarity measures and the optimal values of the MDP. Extensive experiments show… CONTINUE READING
    5 Citations

    Figures and Topics from this paper

    Privacy-preserving spatial keyword location-to-trajectory matching
    An efficient algorithm for spatio-textual location matching
    • PDF
    CAPMAN: Cooling and Active Power Management in big.LITTLE Battery Supported Devices
    • 1


    Metrics for Finite Markov Decision Processes
    • 169
    • Highly Influential
    • PDF
    Measuring the Distance Between Finite Markov Decision Processes
    • 31
    • PDF
    P-Rank: a comprehensive structural similarity measure over information networks
    • 194
    • PDF
    SimRank: a measure of structural-context similarity
    • 1,792
    • Highly Influential
    • PDF
    Scalable and axiomatic ranking of network role similarity
    • 19
    • PDF
    Equivalence notions and model minimization in Markov decision processes
    • 281
    • PDF
    Towards a Unified Theory of State Abstraction for MDPs
    • 286
    • PDF
    SimFusion: measuring similarity using unified relationship matrix
    • 128
    • PDF
    MatchSim: a novel similarity measure based on maximum neighborhood matching
    • 42