Fayyad, U.M., Irani, K.B., 1993. Multi-interval discretization of continuous attribous as preprocessing for classification learning. In: Bajcsy, R. (Ed.), Proc. 13th Internat. Joint Conf. on Artificial Intelligence. Morgan Kau#mann, Los Altos, CA, pp. 1022--1027.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....to pre process the data so that continuous valued attributes are converted to interval valued ones. Determining the cut off points for the intervals can be done in a variety of ways. Currently implemented (or in the working) are algorithms based on Boolean reasoning [14] entropy considerations [4, 5] and 2 statistics [8, 11] ffl Computation of reducts, both in the standard sense as well as object related and approximate ones. As a way to deal with noisy data and to reveal more general patterns, dynamic reducts [2] can be computed. A genetic algorithm [22] for reduct computation is ....

....passing them through a single virtual Apply method on the structure object. For composite structures, the application method may be overloaded to allow for individual processing of the substructures. The return value semantics of the application method are as follows: Handle Algorithm algorithm[5]; Set up the pipeline topology. algorithm[0] ObjectManager: GetAlgorithm(DECISIONTABLEIMPORTER) algorithm[1] ObjectManager: GetAlgorithm(RSESORTHOGONALSCALER) algorithm[2] ObjectManager: GetAlgorithm(RSESGENETICREDUCER) algorithm[3] ....

Fayyad, U., Irani, K. B.: Multi--interval discretization of continuous attributes as preprocessing for classification learning. In: Proc. Thirteenth International Joint Conference on Artificial Intelligence, Morgan Kaufmann (1995) 1022--1027

.... alan icin herhangi bir kesiklestirme algoritmas ile yaplabilir [1, 24] Dikkat edilmesi gereken husus, kesiklestirme uzerinde yaplan calsmalar sembolik ogrenme tarihi kadar eskidir ve bu konudaki yaklasimlar baz on varsaymlar (ya da oncel istatistiki daglmlar) araclgi ile ozel durumlar hedef alr [16, 8]. 7 Eslestirme probleminin daha genel versiyonu veri bagmllk sorgusu ad ile anlr ki bu konu proje kapsam dahilinde ele alnmyacaktr [11] 12 kafesi bizim ihtiyacmz gorur. Bu baglama uygun olan kavram kafesi ornek olarak S ekil 1 de verilmektedir. Kafes icindeki her bir baglant ff R fi ya ....

FAYYAD, U. M., AND IRANI, K. B. Multi interval discretization of continuous attributes for classification learning. In Proceedings of 13th International Joint Conference on Artificial Intelligence (1993), R. Bajcsy, Ed., Morgan Kauffmann, pp. 1022--1027.

....Set up the pipeline topology. algorithm[0] ObjectManager: GetAlgorithm(DECISIONTABLEIMPORTER) algorithm[1] ObjectManager: GetAlgorithm(RSESORTHOGONALSCALER) algorithm[2] ObjectManager: GetAlgorithm(RSESGENETICREDUCER) algorithm[3] ObjectManager: GetAlgorithm(RSESRULEGENERATOR) algorithm[4] = ObjectManager: GetAlgorithm(PROLOGRULEEXPORTER) Set parameters. algorithm[0] SetParameters( filename = c: table.txt ) algorithm[1] SetParameters( mode = save; filename = c: cuts.txt; algorithm[2] SetParameters( seed = 1234; discernibility = normal; algorithm[3] SetParameters( ....

.... Set parameters. algorithm[0] SetParameters( filename = c: table.txt ) algorithm[1] SetParameters( mode = save; filename = c: cuts.txt; algorithm[2] SetParameters( seed = 1234; discernibility = normal; algorithm[3] SetParameters( algorithm[4] SetParameters( filename = c: rules.pl ) Handle Structure structure = Creator: DecisionTable( Processing. for (int i = 0; i 5; i ) structure = structure Apply( algorithm[i] Figure 4: Implementation of pipeline in Figure 3. intermediate structures does not cause memory leaks in the ....

[Article contains additional citation context not shown here]

U. Fayyad, K. B. Irani (1995), Multi-Interval Discretization of Continuous Attributes as Preprocessing for Classification Learning, Proc. Thirteenth International Joint Conference on Artificial Intelligence, Morgan Kaufmann, pp. 1022--1027.

....an induction problem by itself. For most datasets tested here, the real valued attributes were not very useful (glass is a notable exception) In many real world datasets it is probably true that real values are necessary, and the data must undergo a discretization process (Lenarcik Piasta 1992; Fayyad Irani 1993). Acknowledgments The work in this paper was done using the MLC library, partly funded by ONR grant N00014 94 1 0448 and NSF grant IRI9116399. George John has been of great help in this ongoing research. Rob Holte provided us with excellent comments and useful pointers after seeing the first ....

Fayyad, U. M., and Irani, K. B. 1993. Multiinterval discretization of continuous attributes for classification learning. In Bajcsy, R., ed., Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence. Morgan Kaufmann.

.... use of techniques in data mining that draw upon or are based on statistics; namely, in feature selection [12] data dependency involving two variables for constructing data dependency networks [13, 14] classification of objects based on descriptions [7] discretization of continuous values [13, 15], data summarization [14] predicting missing values [16] etc. The motivation behind this trend can be explained by the fact that statistical techniques for data analysis are well developed and in some cases, we do not have any other means to apply. In many data analysis problems statistical ....

.... reduction is related to merging identical tuples following either the substitution of an attribute value by its higher level value in a pre defined generalization hierarchy of categorical values of the attribute [22] or the quantization (or discretization) of continuous (or numeric) values [13, 15, 23]. Vertical reduction is realized by either applying some feature selection methods or using attribute dependency graph [24] We consider vertical reduction as a part of methods for handling redundant data, in Section 3.5. We elaborate on some notable studies on horizontal reduction in the ....

[Article contains additional citation context not shown here]

U. M. Fayyad and K. B. Irani, "Multi interval discretization of continuous attributes for classification learning," in Proceedings of 13th International Joint Conference on Artificial Intelligence (R. Bajcsy, ed.), pp. 1022--1027, Morgan Kauffmann, 1993.

....of the most important characteristics of decision trees and graphs over other hypothesis spaces. We noticed that for continuous features, many splits are made on different thresholds. Since we only merge nodes at the same level, we might consider multi way splits on real features, as suggested in Fayyad Irani (1993). A different approach to feature selection and ordering, which is more feasible in OODGs then in decision trees, is to try different orderings on the features. An ODT on a dataset is defined by a set of tests, one per level, and hence the space of possibilities is much smaller than that of ....

Fayyad, U. M. & Irani, K. B. (1993), Multi-interval discretization of continuous attributes for classification learning, in R. Bajcsy, ed., "Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence", Morgan Kaufmann.

.... use of techniques in data mining that draw upon or are based on statistics; namely, in feature selection [12] data dependency involving two variables for constructing data dependency networks [13, 14] classification of objects based on descriptions [7] discretization of continuous values [13, 15], data summarization [14] predicting missing values [16] etc. The motivation behind this trend can be explained by the fact that statistical techniques for data analysis are well developed and in some cases, we do not have any other means to apply. In many data analysis problems statistical ....

.... reduction is related to merging identical tuples following either the substitution of an attribute value by its higher level value in a pre defined generalization hierarchy of categorical values of the attribute [22] or the quantization (or discretization) of continuous (or numeric) values [13, 15, 23]. Vertical reduction is realized by either applying some feature selection methods or using attribute dependency graph [24] We consider vertical reduction as a part of methods for handling redundant data, in Section 3.5. We elaborate on some notable studies on horizontal reduction in the ....

[Article contains additional citation context not shown here]

U. M. Fayyad and K. B. Irani, "Multi interval discretization of continuous attributes for classification learning," in Proceedings of 13th International Joint Conference on Artificial Intelligence (R. Bajcsy, ed.), pp. 1022--1027, Morgan Kauffmann, 1993.

....the attribute in question. Given this definition JGA builds a language template by allocating one bit for each interval. The above scheme for discretization is however rather rude, assuming that the user can give suitable interval ranges. An alternative method is to apply a discretizer, see e.g. [8, 21], and then treat the discretized intervals as nominal values. 3.5 Expressing disjunctive concepts The language L 1 uses internal disjunctions but doesn t include higher level disjunctions for relating different predicates. Since we are interested in learning multimodal concepts we need to be able ....

U. M. Fayyad and K. B. Irani. Multi-interval discretization of continuous attributes for classification learning. In Proc. of the 13th International Joint Conference on Artifical Intelligence, pages 1022--1027, San Mateo, CA, 1993. Morgan Kaufmann.

No context found.

Fayyad, U.M., Irani, K.B., 1993. Multi-interval discretization of continuous attribous as preprocessing for classification learning. In: Bajcsy, R. (Ed.), Proc. 13th Internat. Joint Conf. on Artificial Intelligence. Morgan Kau#mann, Los Altos, CA, pp. 1022--1027.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC