Breiman, L., J.H. Friedman, R.A. Olshen, and C.J. Stone. Classification and Regression by tree. Wadsworth, Belmont California, 1984. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Delaying the Choice of Bias: A Disjunctive Version Space Approach - Michèle Sebag (1996) (11 citations) (Correct)
....rather than on the number of inconsistencies, in order to ease the interpretation of the results; ranges from 0 to 25 . Parameter M varies in [1: 20] We detail the experiments conducted on two problems: a nominal problem originated from biology and a numerical problem designed by Breiman et al. 1984). Results obtained on other problems are more briefly discussed. 4.2 THE PROMOTER GENE SEQUENCE Examples are composed of sequences of nucleotides. They are described by 59 attributes valued in fA; C; G; Tg. The associated class gives the promoter activity of the example (boolean) The 106 ....
....the description is rather high: no matter what the value of is, the optimal value of M is 8. A naive interpretation would be: it needs at least 8 attributes to make a real difference between two examples . 4. 3 THE WAVEFORM PROBLEM Examples are built by linear combinations of fixed waveforms (Breiman et al. 1984). An example is described by 21 real valued attributes; examples are equally distributed among three classes. The waveform problem presents three difficulties from the standpoint of ML algorithms: classes are overlapping and data involve numerical noise 5 ; last, the separation of the classes is ....
Breiman, L., J.H. Friedman, R.A. Olshen, and C.J. Stone. Classification and Regression by tree. Wadsworth, Belmont California, 1984.
2 nd Order Understandability of Disjunctive Version Spaces - Michèle Sebag (1995) (2 citations) (Correct)
....that building a redundant knowledge can address the predictive accuracy concern. This claim is supported by some first experimental results: we introduce a learner building disjunctive Version Spaces [Mit82] in a logical language; this learner is experimented on a well studied numerical problem [BFOS84]. However, the understandability concern is poorly addressed by redundant learners, because of the size of the produced knowledge. Therefore we define the property of second order understandability, as the ability of a knowledge base to justify any statement inferred from this knowledge base, ....
....of redundant knowledge, via embedding the version space frame into a star like frame. A constraint based approach polynomially achieves the building of such disjunctive version spaces. In section 3, this learner called DIVS (for DIsjunctive Version Space) is experimented on a well studied problem [BFOS84]. Results are discussed and compared to that of some other numerical and logical approaches. Section 4 focuses on the understandability concern; it introduces and discusses the notion of 2 nd order understandability. 1 Goal and State of the Art This section briefly reviews some well known ....
[Article contains additional citation context not shown here]
L. Breiman, J.H. Friedman, R.A. Olshen, and C.J. Stone. Classification and Regression by tree. Wadsworth, Belmont California, 1984.
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