On the handling of fuzziness for continuous-valued attributes in decision tree generation

doi:10.1016/S0165-0114(97)00030-4

Fuzzy Sets and Systems

Volume 99, Issue 3, 1 November 1998, Pages 283-290

https://doi.org/10.1016/S0165-0114(97)00030-4 Get rights and content

Abstract

In this paper, fuzziness existing in the process of generating decision trees by discretizing continuous-valued attributes is considered. In a sense a better way to express this fuzziness via fuzzy numbers is presented using possibility theory. The fact that selection of membership functions in a class of symmetric distributions does not influence the decision tree generation is proved. The validity of using the tree to classify future examples is explained. On the basis of likelihood possibility maximization, the existing algorithm is revised. The revised algorithm leads to more reasonable and more natural decision trees.

References (12)

W. Nather
On possibilistic inference
Fuzzy Sets and Systems
(1990)
W. Zhenyuan et al.
Fuzzy linear regression analysis of fuzzy-valued variables
Fuzzy Sets and Systems
(1990)
J. Cheng et al.
Improved decision trees: a generalized version of ID3
P. Clark et al.
The CN2 induction algorithm
Machine Learning
(1989)
D. Dubois et al.
Possibility Theory
(1988)
U.M. Fayyad et al.
A machine learning algorithm (GID3^∗) for automated knowledge acquisition: improvements and extensions
(1991)

There are more references available in the full text version of this article.

Cited by (54)

Fuzzy nonlinear regression analysis using a random weight network
2016, Information Sciences
Citation Excerpt :
Second, we will theoretically analyze the influence of random initialization on the approximation power of FNRRWN. Third, we will develop the fuzzy RWNs based on granular computing paradigm [28] and maximum uncertainty [36,37,39,41]. Fourth, we will find some practical applications for FNRRWN, e.g., image processing [24,44,55], sentiment analysis for Chinese text [9–11], and protein-protein interaction prediction [46,47].
Modeling a fuzzy-in fuzzy-out system where both inputs and outputs are uncertain is of practical and theoretical importance. Fuzzy nonlinear regression (FNR) is one of the approaches used most widely to model such systems. In this study, we propose the use of a Random Weight Network (RWN) to develop a FNR model called FNR_RWN, where both the inputs and outputs are triangular fuzzy numbers. Unlike existing FNR models based on back-propagation (BP) and radial basis function (RBF) networks, FNR_RWN does not require iterative adjustment of the network weights and biases. Instead, the input layer weights and hidden layer biases of FNR_RWN are selected randomly. The output layer weights for FNR_RWN are calculated analytically based on a derived updating rule, which aims to minimize the integrated squared error between α-cut sets that correspond to the predicted fuzzy outputs and target fuzzy outputs, respectively. In FNR_RWN, the integrated squared error is solved approximately by Riemann integral theory. The experimental results show that the proposed FNR_RWN method can effectively approximate a fuzzy-in fuzzy-out system. FNR_RWN obtains better prediction accuracy in a lower computational time compared with existing FNR models based on BP and RBF networks.
Learning ELM-Tree from big data based on uncertainty reduction
2015, Fuzzy Sets and Systems
Citation Excerpt :
Generally, the scale of a tree is affected by two factors, i.e., the degree of over-partitioning in the induction process, and the selection of heuristic measure for splitting nodes. Information entropy [18,22] and ambiguity [29,34], which respectively reflect the impurity of classes and the uncertainty of the split, are the two most widely used heuristics for splitting nodes during the tree growth. The performances of these two heuristics may differ a lot, however, both of them are uncertainty reduction based methods [31,32] that can be applied to the parallelization of sequential DT.
A challenge in big data classification is the design of highly parallelized learning algorithms. One solution to this problem is applying parallel computation to different components of a learning model. In this paper, we first propose an extreme learning machine tree (ELM-Tree) model based on the heuristics of uncertainty reduction. In the ELM-Tree model, information entropy and ambiguity are used as the uncertainty measures for splitting decision tree (DT) nodes. Besides, in order to resolve the over-partitioning problem in the DT induction, ELMs are embedded as the leaf nodes when the gain ratios of all the available splits are smaller than a given threshold. Then, we apply parallel computation to five components of the ELM-Tree model, which effectively reduces the computational time for big data classification. Experimental studies demonstrate the effectiveness of the proposed method.
Cognitively Inspired Multi-attribute Decision-making Methods Under Uncertainty: a State-of-the-art Survey
2022, Cognitive Computation
Dynamic optimisation based fuzzy association rule mining method
2019, International Journal of Machine Learning and Cybernetics
Predicting algal appearance on mortar surface with ensembles of adaptive neuro fuzzy models: a comparative study of ensemble strategies
2019, International Journal of Machine Learning and Cybernetics
Certain types of soft coverings based rough sets with applications
2019, International Journal of Machine Learning and Cybernetics

View all citing articles on Scopus

View full text

On the handling of fuzziness for continuous-valued attributes in decision tree generation

Abstract

Fuzzy Sets and Systems

Fuzzy Sets and Systems

Improved decision trees: a generalized version of ID3

The CN2 induction algorithm

Machine Learning

Possibility Theory

A machine learning algorithm (GID3∗) for automated knowledge acquisition: improvements and extensions

A machine learning algorithm (GID3^∗) for automated knowledge acquisition: improvements and extensions