Playing Billiards in Version Space

@article{Rujan1997PlayingBI,
  title={Playing Billiards in Version Space},
  author={Pal Rujan},
  journal={Neural Computation},
  year={1997},
  volume={9},
  pages={99-122}
}
A ray-tracing method inspired by ergodic billiards is used to estimate the theoretically best decision rule for a given set of linear separable examples. For randomly distributed examples, the billiard estimate of the single Perceptron with best average generalization probability agrees with known analytic results, while for real-life classification problems, the generalization probability is consistently enhanced when compared to the maximal stability Perceptron. 

References

Publications referenced by this paper.
Showing 1-10 of 30 references

Optimal learning with a neural network

  • T. Watkin
  • Europhysics Letters 21, 871.
  • 1993
Highly Influential
4 Excerpts

Estimation of dependencies from empirical data, SpringerVerlag, Berlin

  • D. Vapnik
  • 1982
Highly Influential
4 Excerpts

Designing pattern categories with extremal paradigm information

  • P. F. Lampert
  • Methodologies of Pattern Recognition, M. S…
  • 1969
Highly Influential
1 Excerpt

Bayesian learning

  • M. Opper
  • Talk at the Neural Networks for Physicists 3…
  • 1993
1 Excerpt

Similar Papers

Loading similar papers…