List of Tables

1. The training set for the Tech Support domain.
2. Information provided by gloves.
3. The training set for the Tech Support domain.
4. Instantiated LoudRun features for the Tech Support domain.
5. Attribution of synthetic features for the Tech Support domain.
6. A general contingency table
7. Random sets of centroids generated for the Tech Support domain.
8. Contingency table for Trial 1.
9. Contingency table for Trial 2.
10. Contingency table for Trial 3.
11. Evaluation of trials 1, 2 and 3 using Gain, Gain Ratio and Chi-Squared disparity measures.
12. Attribution of synthetic features for the Tech Support domain in Trial 2.
13. Numerical attribution of synthetic features for the Tech Support domain in Trial 2.
14. Relative membership (RM) for instantiated features in region 2 of Trial 2.
15. Previously published error rates on the CBF domain.
16. Error rates using the current version of TClass on the CBF domain.
17. Error rates for the naive segmentation algorithm on the CBF domain.
18. Error rates of hidden Markov models on CBF domain.
19. Error rates on the TTest domain.
20. Error rates of naive segmentation on the TTest domain.
21. Error rates when using hidden Markov models on TTest domain.
22. Error rates for high-noise situation (g=0.2) on TTest domain.
23. Error rates with no Gaussian noise (g=0) on TTest domain.
24. Error rates for high-noise situation (g=0.2), but with only 100 examples on TTest domain.
25. Some statistics on the signs used.
26. Error rates for the Nintendo sign language data.
27. Error rates with different TClass parameters.
28. Error rates for the Flock sign language data.
29. Error rates with minor refinements for the Flock data.
30. Class distribution for the ECG learning task.
31. Error rates for the ECG data (dominant beats only)
32. Effect of only allowing ``wide'' maxima and minima on dominant beats. There is not a significant difference in terms of accuracy.
33. Error rates for the ECG data with all beats used

In the Name of The Unique God, the Beneficent, the Merciful