Deogun, J.S.: Feature selection and effective classifiers (1998)
0.00
0.0025282204 = product of:
0.005056441 = sum of:
0.005056441 = product of:
0.010112882 = sum of:
0.010112882 = weight(_text_:a in 2911) [ClassicSimilarity], result of:
0.010112882 = score(doc=2911,freq=10.0), product of:
0.059167966 = queryWeight, product of:
1.153047 = idf(docFreq=37942, maxDocs=44218)
0.051314447 = queryNorm
0.1709182 = fieldWeight in 2911, product of:
3.1622777 = tf(freq=10.0), with freq of:
10.0 = termFreq=10.0
1.153047 = idf(docFreq=37942, maxDocs=44218)
0.046875 = fieldNorm(doc=2911)
0.5 = coord(1/2)
0.5 = coord(1/2)
- Abstract
- Develops and analyzes 4 algorithms for feature selection in the context of rough set methodology. Develops the notion of accuracy of classification that can be used for upper or lower classification methods and defines the feature selection problem. Presents a discussion of upper classifiers and develops 4 features selection heuristics and discusses the family of stepwise backward selection algorithms. Analyzes the worst case time complexity in all algorithms presented. Discusses details of the experiments and results of using a family of stepwise backward selection learning data sets and a duodenal ulcer data set. Includes the experimental setup and results of comparison of lower classifiers and upper classiers on the duodenal ulcer data set. Discusses exteded decision tables
- Footnote
- Contribution to a special issue devoted to knowledge discovery and data mining
- Type
- a