Summary
This chapter is concentrated with the performance characterization of a case-based reasoning (CBR) system. Based on the match score and nonmatch score computed from the cases in the case library, we develop a statistical model for prediction. We estimate the size of a subset of cases, called gallery size, that can generate the optimal error estimate and its confidence on a large population (relative to the size of the gallery). The statistical model is based on a generalized two-dimensional prediction model that combines a hypergeometric probability distribution model with a binomial model explicitly and considers the data distortion problem in large populations. Learning is incorporated in the prediction process in order to find the optimal small gallery size and to improve the prediction performance. During the prediction, the expectation-maximization (EM) algorithm is used to learn the match score and the nonmatch score distributions that are represented as mixture of Gaussians. By learning, the optimal size of small gallery is determined and at the same time the upper bound and the lower bound for the prediction on large populations are obtained. Results are shown using a real-world database with the increasing size of the case library.
Access provided by Autonomous University of Puebla. Download to read the full chapter text
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Kolodner J (1993) Case-based Reasoning. Morgan Kaufmann Publisher.
Perner P, Perner H, Jänichen S (2006) Recognition of airborne fungi spores in digital microscopic images. J Arificial Intelligence in Medicine AIM, Special Issue on CBR, vol. 36, no. 2, pp. 137–157.
Perner P, Jänichen S (2004) Case acquisition and case mining for case-based object recognition. In: Peter Funk, Pedro A. González Calero (Eds.), Advances in Case-Based Reasoning, ECCBR2004, Springer Verlag 2004, vol. 3155, pp. 616–629.
Ming J, Bhanu B (1997) ORACLE: An integrated learning approach for object recognition. J Pattern Recognition and Artificial Intelligence, vol. 11, no. 6, pp. 961–990.
Minor M, Hanft A (1999) Cases with a life-cycle. In: Proc. Intl. Conf. on Case-Based Reasoning Workshops, 1999, pp. 3–8.
Guyon I, Makhoul J (1998) What size test set gives good error rate estimates? J IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 20, no. 1, pp. 52–64.
Mitchell TM (1997) Machine Learning. McGraw Hill.
Duda RO, Hart PE, Stork DG (2000) Pattern Classification. Wiley-Interscience Publication.
Wang R, Bhanu B (2007) Predicting fingerprint biometric performance from a small gallery. J Pattern Recognition Letters, vol. 28, pp. 40–48.
Wayman JL (1999) Error-rate equations for the general biometric system. J IEEE Robotics & Automation Magazine, vol. 6, issue 1, pp. 35–48.
Daugman J (2003) The importance of being random: statistical principles of iris recognition. J Pattern Recognition, vol. 36, no. 2, pp. 279–291.
Phillips PJ, Grother P, Micheals RJ, Blackburn DM, Tabassi E, and Bone M (2003) Face recognition vendor test 2002, Evaluation Report.
Wang R, Bhanu B, Chen H (2005) An integrated prediction model for biometrics. In: Proc. Audio- and Video-based Biometric Person Authentication, New York, pp. 355–364.
Johnson AY, Sun J, Boick AF (2003) Using similarity scores from a small gallery to estimate recognition performance for large galleries. In: Proc. IEEE Int. Workshop on Analysis and Modeling of Faces and Gestures, pp. 100–103.
Grother P, Phillips PJ (2004) Models of large population recognition performance. In: Proc. IEEE Conf. on Computer Vision and Pattern Recognition, vol. 2, pp. 68–75.
Lindenbaum M (1995) Bounds on shape recognition performance. J Pattern Analysis and Machine Intelligence, vol. 17, no. 7, pp. 666–680.
Lindenbaum M (1997) An integrated model for evaluating the amount of data required for reliable recognition. J Pattern Analysis and Machine Intelligence, vol. 19, no. 11, pp. 1251–1264.
Boshra M, Bhanu B (2000) Predicting performance of object recognition. J Pattern Analysis and Machine Intelligence, vol. 22, no. 9, pp. 956–969.
Boshra M, Bhanu B (2001) Predicting an upper bound on SAR ATR performance. J IEEE Trans. on Aerospace and Electronic Systems, vol. 37, no. 3, pp. 876–888.
Wang R, Bhanu B (2005) Learning models for predicting recognition performance. In: Proc. IEEE International Conf. on Computer Vision, vol. 2, pp. 1613–1618.
Tan X, Bhanu B (2002) Robust fingerprint identification. In: Proc. IEEE Int. Conf. on Image Processing, vol. 1, pp. 277–280.
Mood AM, Graybill FA, Boes DC (1974) Introduction to the Theory of Statistics. McGraw,Hill.
Bhanu B, Boshra M, Tan X (2000) Logical templates for feature extraction in fingerprint images. In: Proc. Int. Conf. on Pattern Recognition, vol. 3, pp. 850–854.
Figueiredo MAT, Jain AK (2002) Supervised learning of finite mixture models. J IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 24, no. 3, pp. 381–396.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Bhanu, B., Wang, R. (2008). Learning a Statistical Model for Performance Prediction in Case-Based Reasoning. In: Perner, P. (eds) Case-Based Reasoning on Images and Signals. Studies in Computational Intelligence, vol 73. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-73180-1_7
Download citation
DOI: https://doi.org/10.1007/978-3-540-73180-1_7
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-73178-8
Online ISBN: 978-3-540-73180-1
eBook Packages: EngineeringEngineering (R0)