An Algorithm for Inducing Stochastic Finite Memory Machines


In this paper, we present a novel algorithm for inducing stochastic finite memory machines to match a given example data-set. The motivation for this approach stems for the ANNIE 2006 binary time-series prediction competition. Our approach extends the paradigm of finite memory machines into the stochastic domain, proves the applicability of our new extension to the problem of inducing unknown sequences, and presents some preliminary results on the effectiveness of this approach. Our score on the ANNIE 2006 competition data set was 19606 of 20,000 bits predicted correctly or 98.0349% accuracy. At the time of publication, this was the highest score in the competition. The best result ever on a similar but different sequence was not signficantly different (98.4%) (Ashlock, 2006).

  • Abstract
  • 1 Introduction
  • 2 Grammar Induction
  • 3 The Model
  • 4 Induction
  • 5 Illustrative Example
  • 6 Setting Parameters
  • 7 Results
  • 8 Competition Data Results
  • 9 Conclusion
  • Acknowledgements
  • References

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In