steelplates {rebmix}

Steel Plates Faults Recognition Data

Description

These data are the results of an extraction process from images of faults of steel plates. There are seven different faults: Pastry (1), Z_Scratch (2), K_Scratch (3), Stains (4), Dirtiness (5), Bumps (6), Other faults (7).

Usage

data(steelplates)

Format

steelplates is a data frame with 1941 cases (rows) and 28 variables (columns) named:

1. X_Minimum integer.

2. X_Maximum integer.

3. Y_Minimum integer.

4. Y_Maximum integer.

5. Pixels_Areas integer.

6. X_Perimeter integer.

7. Y_Perimeter integer.

8. Sum_of_Luminosity integer.

9. Minimum_of_Luminosity integer.

10. Maximum_of_Luminosity integer.

11. Length_of_Conveyer integer.

12. TypeOfSteel_A300 binary.

13. TypeOfSteel_A400 binary.

14. Steel_Plate_Thickness integer.

15. Edges_Index continuous.

16. Empty_Index continuous.

17. Square_Index continuous.

18. Outside_X_Index continuous.

19. Edges_X_Index continuous.

20. Edges_Y_Index continuous.

21. Outside_Global_Index continuous.

22. LogOfAreas continuous.

23. Log_X_Index continuous.

24. Log_Y_Index continuous.

25. Orientation_Index continuous.

26. Luminosity_Index continuous.

27. SigmoidOfAreas continuous.

28. Class discrete 1, 2, 3, 4, 5, 6 or 7.

Source

A. Asuncion and D. J. Newman. Uci machine learning repository, 2007. http://archive.ics.uci.edu/ml/.

References

M. Buscema, S. Terzi, W. Tastle. A new meta-classifier. Annual Conference of the North American Fuzzy Information Processing Society - NAFIPS, 2010. doi:10.1109/NAFIPS.2010.5548298.

M. Buscema. MetaNet*: The theory of independent judges. Substance Use & Misuse. 33(2):439-461, 1998. doi:10.3109/10826089809115875.

Examples

## Not run: 
data(steelplates)

# Split dataset into train (75

set.seed(3)

Steelplates <- split(p = 0.75, Dataset = steelplates, class = 28)

# Estimate number of components, component weights and component
# parameters for train subsets.

steelplatesest <- REBMIX(model = "REBMVNORM",
  Dataset = a.train(Steelplates),
  Preprocessing = "histogram",
  cmax = 15,
  Criterion = "BIC")

# Classification.

steelplatescla <- RCLSMIX(model = "RCLSMVNORM",
  x = list(steelplatesest),
  Dataset = a.test(Steelplates),
  Zt = a.Zt(Steelplates))

steelplatescla

summary(steelplatescla)

## End(Not run)