Hybrid Bayesian estimation tree learning with discrete and fuzzy labels

Abstract

Classical decision tree model is one of the classical machine learning models for its simplicity and effectiveness in applications. However, compared to the DT model, probability estimation trees (PETs) give a better estimation on class probability. In order to get a good probability estimation, we usually need large trees which are not desirable with respect to model transparency. Linguistic decision tree (LDT) is a PET model based on label semantics. Fuzzy labels are used for building the tree and each branch is associated with a probability distribution over classes. If there is no overlap between neighboring fuzzy labels, these fuzzy labels then become discrete labels and a LDT with discrete labels becomes a special case of the PET model. In this paper, two hybrid models by combining the naive Bayes classifier and PETs are proposed in order to build a model with good performance without losing too much transparency. The first model uses naive Bayes estimation given a PET, and the second model uses a set of small-sized PETs as estimators by assuming the independence between these trees. Empirical studies on discrete and fuzzy labels show that the first model outperforms the PET model at shallow depth, and the second model is equivalent to the naive Bayes and PET.

Keywords

fuzzy labels / label semantics / random set / probability estimation tree / mass assignment / linguistic decision tree / naive Bayes

References

[1] Quinlan J R. Induction of decision trees.Machine Learning, 1986, 1(1): 81-106

[2] Olaru C, Wehenkel L. A complete fuzzy decision tree technique. Fuzzy Sets and Systems, 2003, 138(2): 221-254

[3] Quinlan J R. C4. 5: programs for machine learning. Morgan Kaufmann, 1993

[4] Baldwin J, Lawry J, Martin T. Mass assignment fuzzy ID3 with applications. In: Proceedings of the Unicom Workshop on Fuzzy Logic: Applications and Future Directions. 1997, 278-294

[5] Janikow C Z. Fuzzy decision trees: issues and methods. IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, 1998, 28(1): 1-14

[6] Huang Z, Gedeon T D, Nikravesh M. Pattern trees induction: a new machine learning method. IEEE Transactions on Fuzzy Systems, 2008, 16(4): 958-970

[7] Qin B, Xia Y, Li F. Dtu: a decision tree for uncertain data. Advances in Knowledge Discovery and Data Mining, 2009: 4-15

[8] Provost F, Domingos P. Tree induction for probability-based ranking. Machine Learning, 2003, 52(3): 199-215

[9] Qin Z, Lawry J. Decision tree learning with fuzzy labels. Information Sciences, 2005, 172(1): 91-129

[10] Qin Z, Lawry J. Prediction trees using linguistic modelling. In: Proceedings of World Congress of International Fuzzy Systems Association (IFSA-05), 2005

[11] Qin Z, Lawry J. Prediction and query evaluation using linguistic decision trees. Applied Soft Computing, 2011, 11(5): 3916-3928

[12] Lawry J. A framework for linguistic modelling. Artificial Intelligence, 2004, 155(1): 1-39

[13] Elkan C. Naive bayesian learning. Technical Report CS97-557, Dept. of Computer Science and Engineering, UCSD, 1997

[14] Blake C, Merz C J. UCI machine learning repository. http://www.ics.uci. edu/∼mlearn/ MLRepository.html

[15] Zadeh L A. Fuzzy logic= computing with words. IEEE Transactions on Fuzzy Systems, 1996, 4(2): 103-111

[16] Zadeh L A. The concept of a linguistic variable and its application to approximate reasoning-I. Information Sciences, 1975, 8(3): 199-249

[17] Sufyan Beg M, Thint M, Qin Z. Pnl-enhanced restricted domain question answering system. In: Proceedings of the 2007 IEEE International Fuzzy Systems Conference. 2007, 1-7

[18] Qin Z, Thint M, Beg M S. Deduction engine design for pnl-based question answering system. In: Proceedings of the 12th International Fuzzy Systems Association World Congress. 2007, 253-262

[19] Lawry J. Modeling and reasoning with vague concepts. Springer, 2006

[20] Lawry J, Shanahan J G, Ralescu A. Modelling with words: learning, fusion, and reasoning within a formal linguistic representation framework. Volume 2873. Springer, 2003

[21] Qin Z, Lawry J. Lfoil: linguistic rule induction in the label semantics framework. Fuzzy Sets and Systems, 2008, 159(4): 435-448

[22] Baldwin J F, Martin T P, Pilsworth B W. Fril-fuzzy and evidential reasoning in artificial intelligence. John Wiley & Sons, Inc., 1995

[23] Zhang W, Qin Z. Dissimilarity measure of logical expressions. In: Proceedings of the 2010 International Conference on Machine Learning and Cybernetics (ICMLC). 2010, 199-203

[24] Zhang W, Qin Z. Clustering data and imprecise concepts. In: Proceedings of the IEEE International Conference on Fuzzy Systems (FUZZ). 2011, 603-608

[25] Jeffrey R C. The logic of decision. University of Chicago Press, 1990

[26] Qin Z, Lawry J. Fuzziness and performance: an empirical study with linguistic decision trees. In: Proceedings of the 12th International Fuzzy Systems Association World Congress on Foundations of Fuzzy Logic and Soft Computing. 2007, 407-416

[27] Randon N J, Lawry J. Classification and query evaluation using modelling with words. Information Sciences, 2006, 176(4): 438-464

[28] Qin Z. Naive bayes classification given probability estimation trees. In: Proceedings of the 5th International Conference on Machine Learning and Applications, ICMLA’06. 2006, 34-42

[29] Qin Z, Lawry J. Hybrid bayesian estimation trees based on label semantics. Lecture Notes in Computer Science, 2005, 896-907

RIGHTS & PERMISSIONS

2014 Higher Education Press and Springer-Verlag Berlin Heidelberg