Towards a computational theory of cognitive maps
Artificial Intelligence
Acquisition of dynamic control knowledge for a robotic manipulator
Proceedings of the seventh international conference (1990) on Machine learning
Using local models to control movement
Advances in neural information processing systems 2
Behavioral synergy without explicit integration
ACM SIGART Bulletin
Enhancing transfer in reinforcement learning by building stochastic models of robot actions
ML92 Proceedings of the ninth international workshop on Machine learning
Case-based reasoning
Topological mapping for mobile robots using a combination of sonar and vision sensing
AAAI'94 Proceedings of the twelfth national conference on Artificial intelligence (vol. 2)
Scaling to domains with irrelevant features
Computational learning theory and natural learning systems: Volume IV
Inductive Logic Programming: Techniques and Applications
Inductive Logic Programming: Techniques and Applications
A study of instance-based algorithms for supervised learning tasks: mathematical, empirical, and psychological evaluations
Spatial Cognition and Computation
The Utility of Global Representations in a Cognitive Map
COSIT 2001 Proceedings of the International Conference on Spatial Information Theory: Foundations of Geographic Information Science
Hi-index | 0.00 |
In this paper we define the task of place learning and describe oneapproach to this problem. Our framework represents distinct places as evidence grids, a probabilistic description of occupancy. Placerecognition relies on nearest neighbor classification, augmented by a registration process to correct for translational differences between the two grids. The learning mechanism is lazy in that it involves the simple storage of inferred evidence grids. Experimental studieswith physical and simulated robots suggest that this approach improves place recognition with experience, that it can handle significant sensornoise, that it benefits from improved quality in stored cases, and thatit scales well to environments with many distinct places. Additionalstudies suggest that using historical information about the robot‘spath through the environment can actually reduce recognition accuracy. Previous researchers have studied evidence grids and place learning,but they have not combined these two powerful concepts, nor have theyused systematic experimentation to evaluate their methods‘ abilities.