Abstract
Traditional learning-from-examples methods assume that examples are given beforehand and all features are measured for each example. However, in many robotic domains the number of features that could be measured is very large, the cost of measuring those features is significant, and thus the robot must judiciously select which features it will measure. Finding a proper tradeoff between theaccuracy (e.g., number of prediction errors) andefficiency (e.g., cost of measuring features) during learning (prior to convergence) is an important part of the problem. Inspired by such robotic domains, this article considers realistic measurement costs of features in the process of incremental learning of classification knowledge. It proposes a unified framework for learning-from-examples methods that trade off accuracy for efficiency during learning, and analyzes two methods (CS-ID3 and CS-IBL) in detail. Moreover, this article illustrates the application of such a cost-sensitive-learning method to a real robot designed for anapproach-recognize task. The resulting robot learns to approach, recognize, and grasp objects on a floor effectively and efficiently. Experimental results show that highly accurate classification procedures can be learned without sacrificing efficiency in the case of both synthetic and real domains.
Article PDF
Similar content being viewed by others
References
Aha, D.W. (1989). Incremental, instance-based learning of independent and graded concept descriptions.Proceedings of the Sixth International Workshop on Machine Learning (pp. 387–391). Ithaca, NY: Morgan Kaufmann.
Aha, D.W., & Gladstone, R.L. (1990). Learning attribute relevance in context in instance-based learning algorithms.Proceedings of the Twelfth Annual Conference of the Cognitive Science Society (pp. 141–148). Cambridge, MA: Lawrence Erlbaum Associates.
Breiman, L., Friedman, J.H., Olsen, R.A., & Stone, C.J. (1984).Classification and regression trees. Belmont, CA: Wadsworth.
Chen, C.H., & Mulgaonkar, P.G. (1990). Feature-utility measures for automatic vision programming.Proceedings of the 1990 IEEE International Conference on Robotics and Automation (pp. 1840–1845). IEEE Computer Society Press.
Cox, L.A. (1989). Incorporating statistical information into expert classification systems to reduce classification costs.Proceedings of the Second International Workshop on Artificial Intelligence and Statistics (pp. 9.1–9.11). Fort Lauderdale, FL:
Dean et al. (1990). Coping with uncertainty in a control system for navigation and exploration.Proceedings of the Eighth National Conference on Artificial Intelligence (pp. 1010–1015). Menlo Park, CA: AAAI Press/The MIT Press.
Feigenbaum, E.A., & Simon, H.A. (1984). EPAM-like models of recognition and learning.Cognitive Science, 8, 305–336.
Fisher, D.H. (1987). Knowledge acquisition via incremental conceptual clustering.Machine Learning, 2, 139–172.
Gennari, J.H. (1989). Focused concept formation.Proceedings of the Sixth International Workshop on Machine Learning (pp. 379–382). Ithaca, NY: Morgan Kaufmann.
Goldberg, K., & Mason, M.T. (1990). Bayesian grasping.Proceedings of the 1990 IEEE International Conference on Robotics and Automation (pp. 1264–1269). IEEE Computer Society Press.
Hong, K.S., Ikeuchi, K., & Gremban, K.D. (1990).Minimum cost aspect classification: a module of a vision algorithm compiler (Technical Report CMU-CS-90-124). School of Computer Science, Carnegie Mellon University, Pittsburgh, PA.
Keller, R.M. (1987). Defining operationality for explanation-based learning.Proceedings of the Fifth National Conference on Artificial Intelligence (pp. 482–487). AAAI Press/The MIT Press.
Laird, J.E., Rosenbloom, P.S., & Newell, A. (1986). Chunking in SOAR: The anatomy of a general learning mechanism.Machine Learning, 1, 11–46.
Mason, M. (1985). Manipulator grasping and pushing operations.Robot hands and the mechanics of manipulation. Cambridge, MA: MIT Press.
Michalski, R.S., & Chilausky, R.L. (1980). Learning by being told and learning from examples: an experimental comparison of the two methods of knowledge acquisition in the context of developing an expert system for soybean disease diagnosis.International Journal of Policy Analysis and Information Systems, 4(2, 125–161.
Miller, D.P. (1989). Execution monitoring for a mobile robot system.Proceedings of the SPIE Conference on Intelligent Control.
Mitchell, T.M. (1977). Version spaces: a candidate elimination approach to rule learning.Proceedings of the Fifth IJCAI (pp. 305–310). Cambridge, MA.
Mitchell, T.M. (1982). Generalization as search.Artificial Intelligence, 18, 203–226.
Mitchell, T.M., Keller, R., & Kedar-Carbelli, S. (1986). Explanation-based generalization: a unifying view.Machine Learning, 1(1, 47–80.
Norton, S.W. (1989). Generating better decision trees.Proceedings of the Eleventh IJCAI (pp. 800–806). Morgan Kaufmann.
Nunez, M. (1991). The use of background knowledge in decision tree induction.Machine Learning, 6(3, 231–250.
Quinlan, J.R. (1986). Induction of decision trees.Machine Learning, 1, 81–106.
Stansfield, S.A. (1989). Reasoning about grasping.Proceedings of the Seventh National Conference on Artificial Intelligence (pp. 768–773). AAAI Press/The MIT Press.
Tan, M. (1991a). Cost-sensitive reinforcement learning for adaptive classification and control.Proceedings of the Ninth National Conference on Artificial Intelligence. AAAI Press/The MIT Press.
Tan, M. (1991b).Cost-sensitive robot learning. Doctoral dissertation, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA. (Technical Report CMU-CS-91-134).
Tan, M., & Schlimmer, J.C. (1989). Cost-sensitive concept learning of sensor use in approach and recognition.Proceedings of the Sixth International Workshop on Machine Learning (pp. 392–295). Ithaca, NY: Morgan Kaufmann (also ref. CMU-CS-89-124).
Tan, M., & Schlimmer, J.C. (1990). Two case studies in cost-sensitive concept acquisition.Proceedings of the Eighth National Conference on Artificial Intelligence (pp. 854–860). Menlo Park, CA: AAAI Press/The MIT Press.
Tan, M., & Schlimmer, J.C. (1991). A cost-sensitive machine learning method for the approach and recognize task.Robotics and Autonomous Systems, 8, 31–45. Elsevier-North Holland.
Trinkle, J., & Paul, R. (1989). The initial grasp liftability chart.IEEE Journal of Robotics and Automation, 5(1) 47–52.
Utgoff, P.E. (1988). ID5: An incremental ID3.Proceedings of the Fifth International Conference on Machine Learning (pp. 107–120). Ann Arbor, MI:
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Tan, M. Cost-sensitive learning of classification knowledge and its applications in robotics. Mach Learn 13, 7–33 (1993). https://doi.org/10.1007/BF00993101
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/BF00993101