Optimized Fuzzy Decision Tree Using Genetic Algorithm

Kim, Myung Won; Ryu, Joung Woo

doi:10.1007/11893295_88

Cited by 5 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 9 shows the results achieved by the different approaches that use GAs, both in training and in test in each data-set. In first place, there are presented the results of the methods for generating FDTs with GAs defined by Kim 35 (GAFDT Kim) and Chang 36 (GAFDT Chang) and then, there are shown the results obtained when applying our methodology. The best global result for each data-set is stressed in bold-face.…”

Section: Study Of the Behaviour Of The Iivfdt Methodsmentioning

confidence: 99%

“…In order to do so, small pattern trees are aggregated to complex ones taking into account the similarity between the tree and the class represented by that tree. • The first approach for optimizing the generation of FDTs that we have selected was defined by Kim and Ryu 35 . In this case, they consider triangular mem-bership functions and they induce the best possible FDT by learning the most suitable fuzzy partition for each variable by means of a GA. To this aim, they apply a classic genetic approach in which they modify the values of the three points which define each triangular membership function.…”

Section: Fuzzy Decision Trees For Comparisonmentioning

confidence: 99%

“…The configuration of the FDTs which are learnt using GAs is introduced in Table 3. In this case, we have used the configuration suggested by Kim and Ryu in 35 , since it provides a solution in a feasible amount of time.…”

Section: Fuzzy Decision Trees For Comparisonmentioning

confidence: 99%

“…Furthermore, we will study the behaviour of the IIVFDT method in comparison with four state-of-the-art FDTs selected in this paper, namely the simple pattern tree 8 , a look-ahead approach 9 , the FDT proposed by Janikow in 6 and a method which fuzzifies the Gini index 7 and also with the C4.5 decision tree 33 , since it is considered a very robust approach in machine learning 34 . We will also compare our new method with two proposals for learning the best FDT by means of GAs, which were defined by Kim and Ryu 35 and Chang et al 36 ; thus, we will show the goodness of our new approach when compared not only with state-of-the-art decision trees, but also with respect to genetically learnt FDTs. Finally, we will study the usefulness of the cooperation between our new approach and the genetic tuning of the lateral position of the membership functions 22 .…”

Section: Introductionmentioning

confidence: 97%

See 3 more Smart Citations

Iivfdt: Ignorance Functions Based Interval-Valued Fuzzy Decision Tree With Genetic Tuning

Sanz

Bustince

Fernández

et al. 2012

Int. J. Unc. Fuzz. Knowl. Based Syst.

View full text Add to dashboard Cite

The choice of membership functions plays an essential role in the success of fuzzy systems. This is a complex problem due to the possible lack of knowledge when assigning punctual values as membership degrees. To face this handicap, we propose a methodology called Ignorance functions based Interval-Valued Fuzzy Decision Tree with genetic tuning, IIVFDT for short, which allows to improve the performance of fuzzy decision trees by taking into account the ignorance degree. This ignorance degree is the result of a weak ignorance function applied to the punctual value set as membership degree. Our IIVFDT proposal is composed of four steps: (1) the base fuzzy decision tree is generated using the fuzzy ID3 algorithm; (2) the linguistic labels are modeled with Interval-Valued Fuzzy Sets. To do so, a new parametrized construction method of Interval-Valued Fuzzy Sets is defined, whose length represents such ignorance degree; (3) the fuzzy reasoning method is extended to work with this representation of the linguistic terms; (4) an evolutionary tuning step is applied for computing the optimal ignorance degree for each Interval-Valued Fuzzy Set. The experimental study shows that the IIVFDT method allows the results provided by the initial fuzzy ID3 with and without Interval-Valued Fuzzy Sets to be outperformed. The suitability of the proposed methodology is shown with respect to both several state-of-the-art fuzzy decision trees and C4.5. Furthermore, we analyze the quality of our approach versus two methods that learn the fuzzy decision tree using genetic algorithms. Finally, we show that a superior performance can be achieved by means of the positive synergy obtained when applying the well known genetic tuning of the lateral position after the application of the IIVFDT method.

show abstract

Section: Study Of the Behaviour Of The Iivfdt Methodsmentioning

confidence: 99%

Section: Fuzzy Decision Trees For Comparisonmentioning

confidence: 99%

Section: Fuzzy Decision Trees For Comparisonmentioning

confidence: 99%