On shape of pattern error function, initializations and intrinsic dimensionality in ann classifier design

Informatica (Netherlands)

Volumn 4, Issue 3-4, 1993, Pages 360-383

  Raudys, Šarūnas ^a  

^a VILNIUS UNIVERSITY   (Lithuania)

Author keywords

Feed forward neural nets; Generalization; Initialization; Insufficient learning; Intrinsic dimensionality; Training sample size

Indexed keywords

EID: 84974762231     PISSN: 08684952     EISSN: None     Source Type: Journal    
DOI: 10.3233/INF-1993-43-410     Document Type: Article

References (18)

1. Baum, E.B. (1990). When k-nearest neighbor and back propagation accurate for feasible sized sets of examples. In L.B. Almeida and C.J. Wellekens (Eds.), Proceedings of the EURASP Workshop on Neural Networks, Pattern Recognition Group, Department of Applied Physics, Delft University of Technology, Delft.
2. Deev, A.D. (1970). Representation of statistics of discriminant analysis and asymptotic expansions in dimensionalities comparable with sample size. Reports of Academy of Sciences of the USSR, 195(4), 756-762 (in Russian).
3. Duin, R.P.W. (1993). Superlearning capabilities of neural networks. In Proc. of the 8th Scandinavian Conference on Image Analysis, NOVIM, Norwegian Society for Image Processing and Pattern Recognition, Tromso, Norway, pp. 547-554.
4. Jain, A., and Š. Raudys (1992). On training sample size and complexity of artificial neural net classifier. Information, 3(2), 301-337.
5. Koford, J.S., and G.F. Groner (1966). The use of an adaptive threshold element to design a linear optimal pattern classifier. IEEE Trans. Inf. Theory, IT-12, 42-50.
6. Korn, G., and T. Korn (1961). Mathematical Handbook for Scientists and Engineers. MGraw-Hiil, NY.
7. Kraaijveld, M.A. (1993). Small sample behavior of multi-lager feedforward network classifiers: theoretical and practical aspects. Ph. D. Thesis, Pattern Recognition Group, Department of Applied Physics, Delft University of Technology, Delft.
8. Pikelis, V. (1976). Comparison of methods of computing the expected classification errors. Automation and Remote Control, 5, 59-63 (in Russian).
9. Raudys, Š., and A.K. Jain (1991). Small sample size effects in statistical pattern recognition: Recommendations for practitioners. IEEE Trans, on PatternAnalysis and Machine Intelligence, 13, 252-264.
10. Raudys, Š., and A. Jain (1991a). Small sample problems in designing artificial neural, networks. In I.K. Sethi and A.K. Jain (Eds.), Elsevier Sci. Publ.NY, pp. 33-50.
11. Raudys, Š., and V. Pikelis (1980). On dimensionality, sample size, classification error and complexity of classification algorithm in pattern recognition. IEEETrans, on PA MI, PAMI-2(3), 242-252.
12. Raudys, Š. (1991). On the effectiveness of Parzen window classifier. Information, 2(3), 434-454.
13. Raudys, Š. (1992). On amount at a prior information in designing the classification algorithm. Proc. Acad, of Sciences of the USSR. Tech. Cyber., 4, 168-174.
14. Rumelhart, D.E., G.E. Hinton and R.J. Williams (1986). Learning internal representations by error propagation. In D.E. Rumelhart and J.L. KcClelland (Eds.), Parallel Distributed Processing, MA:MIT Pres, Cambridge.
15. Schmidt, W.F. (1993). Neural pattern classifying systems, theory and experiments with trainable pattern classifiers. Ph. D. Thesis, Pattern Recognition Group, Department of Applied Physics, Delft University of Technology, Delft.
16. Vapnik, V.N. (1979). Estimation of Dependences Based on Empirical Data. Springer, NY.
17. Wyman, F., D. Young and D. Turner (1990). A comparison of asymptotic error rate expansions for the sample linear discriminant function. Pattern Recognition, 23, 775-783.
18. Zoritch, V. (1984). Mathematical Analysis. Part 2. Nauka, Moscow.