Small data sets for binary supervised classification:
The table below contains data sets used in the joint project of the University of Cologne and the Hochschule Merseburg “Classifying real-world data with the DDα-procedure”. Comprehensive description of the methodology, and experimental settings and results of the study are presented in the work (please cite if you find this useful):
Mozharovskyi, P., Mosler, K., and Lange, T. (2015): Classifying real-world data with the DDα-procedure. Advances in Data Analysis and Classification, 9(3), 287–314. [arXiv:1407.5185]
50 binary classification tasks have been obtained from partitioning 33 freely accessible data sets. Multiclass problems were reasonably split into binary classification problems, some of the data set were slightly processed by removing objects or attributes and selecting prevailing classes. Each data set is provided with a (short) description and brief descriptive statistics. The name reflects the origination of the data. A letter after the name is a property filter, letters (also their combinations) in brackets separated by “vs” are the classes opposed. The letters (combinations or words) stand for labels of classes (names of properties) and are intuitive. Each description contains a link to the original data.
The data have been collected as open source data in January 2013. The owner of this web page decline any responsibility regarding their correctness or consequences of their usage. If you publish material based on these data, please quote the original source. Special requests regarding citations are found on data set’s web page.
Download all the data sets as a single *.zip: zipAll
Data table:
| # | Dataset | n1 | n2 | n1+n2 | d | ln(n1/n2) | (n1+n2)/d | ties | Download |
|---|---|---|---|---|---|---|---|---|---|
| 1. | Baby | 161 | 86 | 247 | 5 | 0,626 | 49,4 | 0 | dat zip |
| 2. | Banknoten | 100 | 100 | 200 | 6 | 0 | 33,3 | 0 | dat zip |
| 3. | Biomedical | 67 | 127 | 194 | 4 | -0,635 | 48,5 | 0 | dat zip |
| 4. | Blood Transfusion | 178 | 570 | 748 | 3 | -1,171 | 249,3 | 246 | dat zip |
| 5. | Breast Cancer Wisconsin | 458 | 241 | 699 | 9 | 0,642 | 77,7 | 236 | dat zip |
| 6. | Bupa Liver Disorder | 145 | 200 | 345 | 6 | -0,329 | 57,5 | 4 | dat zip |
| 7. | Chemical Diabetes (C vs N) | 36 | 76 | 112 | 5 | -0,755 | 22,4 | 0 | dat zip |
| 8. | Chemical Diabetes (C vs O) | 36 | 33 | 69 | 5 | 0,086 | 13,8 | 0 | dat zip |
| 9. | Chemical Diabetes (N vs O) | 76 | 33 | 109 | 5 | 0,833 | 21,8 | 0 | dat zip |
| 10. | Cloud | 54 | 54 | 108 | 7 | 0 | 15,4 | 0 | dat zip |
| 11. | Crabs (B vs O) | 100 | 100 | 200 | 5 | 0 | 40,0 | 0 | dat zip |
| 12. | Crabs (M vs F) | 100 | 100 | 200 | 5 | 0 | 40,0 | 0 | dat zip |
| 13. | Crabs B (M vs F) | 50 | 50 | 100 | 5 | 0 | 20,0 | 0 | dat zip |
| 14. | Crabs F (B vs O) | 50 | 50 | 100 | 5 | 0 | 20,0 | 0 | dat zip |
| 15. | Crabs M (B vs O) | 50 | 50 | 100 | 5 | 0 | 20,0 | 0 | dat zip |
| 16. | Crabs O (M vs F) | 50 | 50 | 100 | 5 | 0 | 20,0 | 0 | dat zip |
| 17. | Cricket (C vs P) | 78 | 78 | 156 | 4 | 0 | 39,0 | 7 | dat zip |
| 18. | Diabetes (of Pima Indians) | 268 | 500 | 768 | 8 | -0,616 | 96,0 | 0 | dat zip |
| 19. | Ecoli (CP vs IM) | 143 | 77 | 220 | 5 | 0,621 | 44,0 | 0 | dat zip |
| 20. | Ecoli (CP vs PP) | 143 | 52 | 195 | 5 | 1,012 | 39,0 | 0 | dat zip |
| 21. | Ecoli (IM vs PP) | 77 | 52 | 129 | 5 | 0,392 | 25,8 | 0 | dat zip |
| 22. | Gemsen (M vs F) | 796 | 553 | 1349 | 6 | 0,365 | 224,8 | 27 | dat zip |
| 23. | Glass (F vs NF) | 70 | 76 | 146 | 9 | -0,083 | 16,2 | 1 | dat zip |
| 24. | Groessen (M vs F) | 116 | 114 | 230 | 3 | 0,020 | 76,7 | 0 | dat zip |
| 25. | Haberman’s Survival | 225 | 81 | 306 | 3 | 1,022 | 102,0 | 23 | dat zip |
| 26. | Heart | 120 | 150 | 270 | 13 | -0,223 | 20,8 | 0 | dat zip |
| 27. | Hemophilia | 30 | 45 | 75 | 2 | -0,400 | 37,5 | 0 | dat zip |
| 28. | Indian Liver Patient (1 vs 2) | 414 | 165 | 579 | 10 | 0,920 | 57,9 | 13 | dat zip |
| 29. | Indian Liver Patient (M vs F) | 140 | 439 | 579 | 9 | -1,139 | 64,3 | 13 | dat zip |
| 30. | Iris Plants (SET vs VER) | 50 | 50 | 100 | 4 | 0 | 25,0 | 2 | dat zip |
| 31. | Iris Plants (SET vs VIR) | 50 | 50 | 100 | 4 | 0 | 25,0 | 3 | dat zip |
| 32. | Iris Plants (VER vs VIR) | 50 | 50 | 100 | 4 | 0 | 25,0 | 1 | dat zip |
| 33. | Irish Educational Transitions (M vs F) | 250 | 250 | 500 | 5 | 0 | 100,0 | 44 | dat zip |
| 34. | Kidney (M vs F) | 20 | 56 | 76 | 5 | -1,022 | 15,2 | 0 | dat zip |
| 35. | PIMA (training) | 132 | 68 | 200 | 7 | 0,663 | 28,6 | 0 | dat zip |
| 36. | Plasma Retinol and Beta-Carotene Levels (M vs F) | 273 | 42 | 315 | 13 | 1,872 | 24,2 | 0 | dat zip |
| 37. | Segmentation (C vs W) | 330 | 330 | 660 | 10 | 0 | 66,0 | 62 | dat zip |
| 38. | Social Mobility (I vs NI) | 578 | 578 | 1156 | 5 | 0 | 231,2 | 45 | dat zip |
| 39. | Social Mobility (W vs B) | 578 | 578 | 1156 | 5 | 0 | 231,2 | 8 | dat zip |
| 40. | Teaching Assistan Evaluation (E vs NE) | 29 | 122 | 151 | 5 | -1,427 | 30,2 | 43 | dat zip |
| 41. | Tennis (M vs F) | 42 | 45 | 87 | 15 | -0,073 | 5,8 | 0 | dat zip |
| 42. | Tips (D vs N) | 176 | 68 | 244 | 6 | 0,952 | 40,7 | 1 | dat zip |
| 43. | Tips (M vs F) | 87 | 157 | 244 | 6 | -0,598 | 40,7 | 1 | dat zip |
| 44. | US Crime (S vs N) | 16 | 31 | 47 | 13 | -0,654 | 3,6 | 0 | dat zip |
| 45. | Vertebral Column | 210 | 100 | 310 | 6 | 0,742 | 51,7 | 0 | dat zip |
| 46. | Veteran Lung Cancer (S vs T) | 69 | 68 | 137 | 7 | 0,010 | 19,6 | 0 | dat zip |
| 47. | Vowel (M vs F) | 528 | 462 | 990 | 13 | 0,131 | 76,2 | 0 | dat zip |
| 48. | Wine (1 vs 2) | 59 | 71 | 130 | 13 | -0,186 | 10,0 | 0 | dat zip |
| 49. | Wine (1 vs 3) | 59 | 48 | 107 | 13 | 0,207 | 8,2 | 0 | dat zip |
| 50. | Wine (2 vs 3) | 71 | 48 | 119 | 13 | 0,392 | 9,2 | 0 | dat zip |
Links to other web-pages containing data sets:
- http://archive.ics.uci.edu/ml [UC Irvine Machine Learning Repository]
- http://cran.r-project.org/web/packages
- http://lib.stat.cmu.edu/datasets [StatLib data sets of the Department of Statistics at Carnegie Mellon University]
- http://stat.ethz.ch/Teaching/Datasets [Teaching data sets of the ETH Zürich]
- http://www.stats.ox.ac.uk/pub/PRNN [Data sets for the book Pattern Recognition and Neural Networks by B.D. Ripley]
- https://www.dbs.ifi.lmu.de/research/outlier-evaluation/DAMI [Data sets for outlier/anomaly detection]
- https://paperswithcode.com/datasets [Data sets from ”papers with code” (for a number of tasks and modalities)]