Adaptive state space partitioning for reinforcement learning Export

@article{citeulike:1723595, abstract = {The convergence property of reinforcement learning has been extensively investigated in the field of machine learning, however, its applications to real-world problems are still constrained due to its computational complexity. A novel algorithm to improve the applicability and efficacy of reinforcement learning algorithms via adaptive state space partitioning is presented. The proposed temporal difference learning with adaptive vector quantization (TD-AVQ) is an online algorithm and does not assume any a priori knowledge with respect to the learning task and environment. It utilizes the information generated from the reinforcement learning algorithms. Therefore, no additional computations on the decisions of how to partition a particular state space are required. A series of simulations are provided to demonstrate the practical values and performance of the proposed algorithms in solving robot motion planning problems.}, author = {Lee, Ivan S. and Lau, Henry Y.}, citeulike-article-id = {1723595}, citeulike-linkout-0 = {http://dx.doi.org/10.1016/j.engappai.2004.08.005}, citeulike-linkout-1 = {http://www.sciencedirect.com/science/article/B6V2M-4DFT21W-1/2/9dcaea7041a89f608febeac72cc66999}, doi = {10.1016/j.engappai.2004.08.005}, journal = {Engineering Applications of Artificial Intelligence}, keywords = {adaptive, continuous\_space, discretization, partitioning}, month = {September}, number = {6}, pages = {577--588}, posted-at = {2007-10-03 11:34:32}, priority = {0}, title = {Adaptive state space partitioning for reinforcement learning}, url = {http://dx.doi.org/10.1016/j.engappai.2004.08.005}, volume = {17}, year = {2004} }

TY - JOUR ID - citeulike:1723595 L3 - citeulike-article-id:1723595 N2 - The convergence property of reinforcement learning has been extensively investigated in the field of machine learning, however, its applications to real-world problems are still constrained due to its computational complexity. A novel algorithm to improve the applicability and efficacy of reinforcement learning algorithms via adaptive state space partitioning is presented. The proposed temporal difference learning with adaptive vector quantization (TD-AVQ) is an online algorithm and does not assume any a priori knowledge with respect to the learning task and environment. It utilizes the information generated from the reinforcement learning algorithms. Therefore, no additional computations on the decisions of how to partition a particular state space are required. A series of simulations are provided to demonstrate the practical values and performance of the proposed algorithms in solving robot motion planning problems. IS - 6 JF - Engineering Applications of Artificial Intelligence EP - 588 TI - Adaptive state space partitioning for reinforcement learning VL - 17 SP - 577 KW - adaptive KW - continuous_space KW - discretization KW - partitioning AU - Lee, Ivan AU - Lau, Henry PY - 2004/09// UR - http://dx.doi.org/10.1016/j.engappai.2004.08.005 ER -