P. Cucka, N. S. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. International Journal of Pattern Recognition and Artificial Intelligence, 10(5):429--446, 1996.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the goal has been found. Our problem, on the other hand, is to find the shortest path to the goal node for future usage. The search continues until the best path to the goal node has been found. Next, we describe briefly some of the work done on navigation in partially known graphs. Cucka et al. [4] have introduced navigation algorithms for sensory based environments such as automated robots moving in a room. They have used Depth First Search (DFS) based navigation algorithms, choosing the next node that sary, thus making the graph of the network practically fully known. Our algorithm may be ....

....navigations which are presented below. P DFS Positional DFS: This DFS based navigation algorithm sorts the neighbors according to their Euclidean distance from the target node, choosing to try the node with minimum distance to the target node first. This variation was first introduced in [4]. D DFS Directional DFS: This DFS based navigation algorithm sorts the neighbors according to the direction of the edge between them and the current node. It first chooses the node with the smallest difference in angle between the line from that node to the current node, and the line from the ....

[Article contains additional citation context not shown here]

P. Cucka, N. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. IJPRAI, 10:429--446, 1996.

....triangulations. They give a randomized oblivious routing algorithm that works for any triangulation, and ask whether there is a deterministic oblivious routing algorithm for all triangulations. They also give a competitive O(1) memory routing algorithm for Delaunay triangulations. Cucka et al. [5] experimentally evaluate the performance of routing algorithms very similar to those described by Kranakis et al. 8] and Bose and Morin [4] When considering the Euclidean distance travelled during point to point routing, their results show that the GREEDY routing algorithm [4] performs better ....

P. Cucka, N. S. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. International Journal of Pattern Recognition and Artificial Intelligence, 10(5):429--446, 1996.

....function. Improved efficiency in real time search has been obtained by restricting the use of heuristic exploitation [28] and by not requiring the algorithm to find an optimal policy [19] In more specific problem settings, more informed heuristics may be applied, for example Cucka et al. [12] use information about the geometric direction of the goal to heuristically improve a depth first search exploration process. Methods for solving Markov decision processes (MDPs) also involve an interleaved tradeoff between exploration and exploitation [26] In these problems, the environment is ....

P. Cucka, N. S. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. International Journal of Pattern Recognition and Artificial Intelligence, 10(5), 1996.

.... by their biological activity, vibration control, geographic information systems, neighbors of the origin in the R 8 lattice, stress analysis, stability of robot grasps [Belsis et al. 1995] spectrometry [Boardman 1993] constrained control allocation [Bordignon and Durham 1995] robot navigation [Cucka et al. 1995], micromagnetic modeling [Porter et al. 1996] and invariant sets of deltasigma modulators [Zhang et al. 1994] ACKNOWLEDGMENTS Albert Marden and Victor Milenkovic provided excellent environments for completing this work. The referees comments greatly improved the presentation and content of ....

Cucka, P., Netanyahu, N., and Rosenfeld, A. 1995. Learning in navigation: Goal finding in graphs. Technical Report CAR-TR-759, Center for Automation Research, University of Maryland.

....function. Improved efficiency in real time search has been obtained by restricting the use of heuristic exploitation [27] and by not requiring the algorithm to find an optimal policy [18] In more specific problem settings, more informed heuristics may be applied, for example Cucka et al. [9] use information about the geometric direction of the goal to heuristically improve the exploration process. In this paper, we compare on line and off line exploration for a repeated task, where the agent is given some particular task(s) to perform some number of times. We assume that the agent ....

....agent uses the best known path to get to its current target (Step 7a) 3.1.1 Discovery and expected cost of default paths In order to traverse an exploration edge between two known nodes v 1 and v 2 , we apply a depth first search strategy to find v 2 starting from v 1 . Following Cucka et al. [9], we use the minimal angle heuristic to inform the search. 5 Note that there may be an exploration edge between nodes v 1 and v 2 , even if there is no real edge in the environment between them. An exploration edge merely denoted the possibility of exploring the territory between the nodes ....

[Article contains additional citation context not shown here]

P. Cucka, N. S. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. International Journal of Pattern Recognition and Artificial Intelligence, 10(5), 1996.

.... manufacturing [1] classification of molecules by their biological activity, vibration control, geographic information systems, neighbors of the origin in the R 8 lattice, stress analysis, stability of robot grasps [5] spectrometry [6] constrained control allocation [8] robot navigation [17], micromagnetic modeling [38] and invariant sets of delta sigma modulators [45] Acknowledgments: A special thanks to Albert Marden and Victor Milenkovic for providing excellent environments for completing this work. The referees comments greatly improved the presentation and content of this ....

P. Cucka, N.S. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. Technical Report CAR-TR-759, Center for Automation Research, University of Maryland, 1995.

....efficient algorithms to search for the desired paths. If the agent is able to do this, we can say that it has learned to navigate , in the sense that it has learned something about the space in which it is navigating and can use this information to improve its navigational efficiency. In [Cucka et al. 1995], we studied this concept of learning to navigate using a goal finding task in a discrete space, represented by a graph embedded in the plane. We assumed that the agent always knows its position in the plane, as well as the position of the goal; but since it does not know the structure of the ....

Cucka, P., Netanyahu, N.S., and Rosenfeld, A. "Learning in navigation: goal finding in graphs", Center for Automation Research Technical Report CAR-TR-759, University of Maryland, College Park, MD, 1995.

No context found.

P. Cucka, N. S. Netanyahu, and A. Rosenfeld. Learning in navigation: Goal finding in graphs. International Journal of Pattern Recognition and Artificial Intelligence, 10(5):429--446, 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC