User Tools

Site Tools

Sidebar

  1. Overview
  2. Applications
  3. Quick start
  4. Datasets
  5. Patterns
  6. Pattern development
  7. Algorithms
  8. Configuration
  9. Best practice
  10. FAQ
  11. Tricks
  12. Process
  13. Contact
pattern

Patterns

Patterns describe construction of a predictor. For example sum pattern, which sums column by subject in table could be described as: 

SELECT subject
     , SUM(@COLUMN) AS predictor
FROM @TABLE
GROUP BY subject;

List of patterns

  1. Absolute change
  2. Aggregated differences
  3. Aggregated ratios
  4. Aggregated time
  5. Child concentration
  6. Child snapshot
  7. Children statistics
  8. Cluster
  9. Count of changes
  10. Count of loops
  11. Direct field
  12. Direct field is equal to
  13. Direct field snapshot
  14. Distance to cluster seed
  15. Integral (AUC)
  16. Interval aggregation
  17. Interval ratio
  18. Measure of cyclic behavior
  19. Number of child types
  20. Number of children
  21. Number of particular children
  22. Parent field
  23. Parent from lookup
  24. Predictor decoration
  25. Ratio
  26. Ratio of aggregates/frequencies
  27. Ratio on child
  28. Regularity (R2, KS)
  29. Relative change (CAGR)
  30. Segment comparison
  31. Simple loop
  32. Slope
  33. Time of peak
  34. Time since
  35. Time since on child
  36. Weighted aggregation
  37. Zero crossing

Page Tools