Filters








3,464 Hits in 3.3 sec

IMPROVING STABILITY OF DECISION TREES

MARK LAST, ODED MAIMON, EINAT MINKOV
2002 International journal of pattern recognition and artificial intelligence  
on decision tree learning (Breiman et al., 1984 and Quinlan, 1993) , existing methods of constructing decision trees from data suffer from a major problem of instability.  ...  If an algorithm is unstable, the cross-validation results become estimators with high variance (Liu and Motoda, 1998), which means that an Last, Maimon, Minkov: Improving Stability of Decision Trees  ...  The CID3 algorithm incrementally generates a multi-layer network, where each hidden layer is associated with a decision tree grown by the ID3 algorithm (Quinlan, 1986 The idea of using a restricted set  ... 
doi:10.1142/s0218001402001599 fatcat:t7dvin6fn5dgjawijzlusrmy2y

An Improved Algorithm for Incremental Induction of Decision Trees [chapter]

Paul E. Utgoff
1994 Machine Learning Proceedings 1994  
The ID3 algorithm and its variants are compared in terms of theoretical complexity and empirical behavior.  ...  , which can result in smaller decision trees.  ...  Incremental Induction of Decision Trees ID3 is a useful concept-learning algorithm because it can efficiently construct a decision tree that generalizes well.  ... 
doi:10.1016/b978-1-55860-335-6.50046-5 dblp:conf/icml/Utgoff94 fatcat:wlzehuz7uvdahii5huz72fh2my

Interruptible anytime algorithms for iterative improvement of decision trees

Saher Esmeir, Shaul Markovitch
2005 Proceedings of the 1st international workshop on Utility-based data mining - UBDM '05  
Therefore, most of the existing algorithms for decision tree induction use a greedy approach based on local heuristics.  ...  Finding a minimal decision tree consistent with the examples is an NP-complete problem.  ...  Therefor, in order to produce a better estimation of the tree size, instead of calling ID3 once, LSID3 samples the space of "good" trees by repeatedly invoking a stochastic version of ID3 (SID3).  ... 
doi:10.1145/1089827.1089837 fatcat:bajalldksjbttgfe32kapxjktm

Improvement of Data Stream Decision Trees

Sarah Nait Bahloul, Oussama Abderrahim, Aya Ichrak Benhadj Amar, Mohammed Yacine Bouhedadja
2022 International Journal of Data Warehousing and Mining  
Hoeffding Tree is a method to, incrementally, build decision trees. Since its proposition in the literature, it has become one of the most popular tools of data stream classification.  ...  Several improvements have since emerged. Hoeffding Anytime Tree was recently introduced and is considered one of the most promising algorithms.  ...  Decision Trees A decision tree is a predictive model used to represent classification and regression (Oded and Lior, 2014) .  ... 
doi:10.4018/ijdwm.290889 fatcat:kofchkkz4fcvzdar3y52padqjm

Clus-DTI: improving decision-tree classification with a clustering-based decision-tree induction algorithm

Rodrigo C. Barros, Márcio P. Basgalupp, André C. P. L. F. de Carvalho, Marcos G. Quiles
2012 Journal of the Brazilian Computer Society  
Our intention is to investigate how clustering data as a part of the induction process affects the accuracy and complexity of the generated models.  ...  Decision-tree induction is a well-known technique for assigning objects to categories in a white-box fashion.  ...  Alex A. Freitas for his valuable comments, which helped improving this paper.  ... 
doi:10.1007/s13173-012-0075-5 fatcat:47erso2hgvdxrhdpnz6nbpto4e

Improving learning accuracy of fuzzy decision trees by hybrid neural networks

E.C.C. Tsang, X.Z. Wang, D.S. Yeung
2000 IEEE transactions on fuzzy systems  
This paper proposes using a hybrid neural network to improve the learning accuracy of Fuzzy ID3 algorithm which is a popular and powerful method of fuzzy rule extraction without much computational effort  ...  The synergy between fuzzy decision tree induction and hybrid neural network offers new insight into the construction of hybrid intelligent systems.  ...  FUZZY ID3 ALGORITHM One popular and powerful heuristic method for generating crisp decision trees is called ID3.  ... 
doi:10.1109/91.873583 fatcat:gkdcjgyccvcz5psrirdzkymtfe

ConfDTree: Improving Decision Trees Using Confidence Intervals

Gilad Katz, Asaf Shabtai, Lior Rokach, Nir Ofek
2012 2012 IEEE 12th International Conference on Data Mining  
The experimental study indicates that the proposed post-processing method consistently and significantly improves the predictive performance of decision trees, particularly for small, imbalanced or multi-class  ...  In this paper we present ConfDTree -a post-processing method which enables decision trees to better classify outlier instances.  ...  For the binary datasets both the ConfDTree COMB and ConfDTree NORM outperformed the original version of the decision tree.  ... 
doi:10.1109/icdm.2012.19 dblp:conf/icdm/KatzSRO12 fatcat:z6apzb4lb5axtidqaptzu4epza

An Improved Collaborative Pruning Using Ant Colony Optimization and Pessimistic Technique of C5.0 Decision Tree Algorithm

I. B. Ayinla, S. O. Akinola
2020 Zenodo  
This paper presents a collaborative pruning model to improve on the classification efficiency of DTs. The model generates two forests using gain ratio: virgin and pessimistic pruned forest.  ...  Besides, it had a better accuracy of 0.98% than the closest C5.0 DT algorithm, especially in a relatively big dataset. Keywords—Decision Tree, Ant Colony, Pheromone trails, Rule-based, Forest  ...  A technique was then introduced known as pruning, which has potential to improve generalization of the decisions trees and reduce the size.  ... 
doi:10.5281/zenodo.4427699 fatcat:5rod3hcr4rh7tki7qgkrthjs2a

Theoretical Study of Decision Tree Algorithms to Identify Pivotal Factors for Performance Improvement: A Review

Pooja Gulati, Amita Sharma, Manish Gupta
2016 International Journal of Computer Applications  
A variety of decision tree algorithms are proposed in the literature like ID3 (Iterative Dichotomiser 3), C4.5 (successor of ID3), CART (Classification and Regression tree), CHAID (Chi-squared Automatic  ...  Decision tree is a data mining technique used for the classification and forecasting of the data.  ...  That's why there is a continuous enhancement needed in the field of decision tree generation.  ... 
doi:10.5120/ijca2016909926 fatcat:z3hcbr66inaalay2flia2obfyi

FDT 2.0: Improving scalability of the fuzzy decision tree induction tool - integrating database storage

Erin-Elizabeth A. Durham, Xiaxia Yu, Robert W. Harrison
2014 2014 IEEE Symposium on Computational Intelligence in Healthcare and e-health (CICARE)  
The freeware fuzzy decision tree induction tool, FDT, is a scalable supervised-classification software tool implementing fuzzy decision trees. It is based on an optimized fuzzy ID3 (FID3) algorithm.  ...  FDT 2.0 improves upon FDT 1.0 by bridging the gap between data science and data engineering: it combines a robust decisioning tool with data retention for future decisions, so that the tool does not need  ...  Acknowledgments This work was supported in part by the Georgia Cancer Coalition (RWH is a Georgia Cancer Scholar) and the Georgia State University Molecular Basis of Disease Initiative.  ... 
doi:10.1109/cicare.2014.7007853 pmid:29226916 pmcid:PMC5721675 dblp:conf/cicare/DurhamYH14 fatcat:hb7r27qq6rdapguiga7geh5tnm

Real Time and Offline Network Intrusion Detection using Improved Decision Tree Algorithm

G. SunilKumar
2012 International Journal of Computer Applications  
Experimental result shows this improved decision tree classifier gives effective decision rules compare to existing decision tree techniques like ID3 and C45 algorithms.  ...  In this paper improved, decision tree is implemented in order to detect network attacks like TCP SYN , Ping of Death, ARP Spoof attacks.  ...  A decision tree can be expressed as a recursive partition of the instance data space.  ... 
doi:10.5120/7541-0482 fatcat:z4gmzy3pmfbohnkzypy75lvhiu

Improve Decision Trees for Probability-Based Ranking by Lazy Learners

Han Liang, Yuhong Yan
2006 Proceedings - International Conference on Tools with Artificial Intelligence, TAI  
This paper aims to improve the ranking performance under decision-tree paradigms by presenting two new models.  ...  Existing work shows that classic decision trees have inherent deficiencies in obtaining a good probability-based ranking (e.g. AUC).  ...  The final version is C4.4. They also pointed out that bagging, an ensemble method, could greatly improve decision trees in terms of probability-based ranking. Ferri et al.  ... 
doi:10.1109/ictai.2006.65 dblp:conf/ictai/LiangY06 fatcat:binl4buyk5dmtgjwcwbpibcx5m

Extracting Useful Rules Through Improved Decision Tree Induction Using Information Entropy

Mohd. Mahmood Ali, Mohd. S. Qaseem, Lakshmi Rajamani, Govardhan A
2013 International Journal of Information Sciences and Techniques  
We suggest improvements to the existing C4.5 decision tree algorithm.  ...  Modified DMQL queries are used to understand and explore the shortcomings of the decision trees generated by C4.5 classifier for education dataset and the results are compared with the proposed approach  ...  C4.5 classifier [1] , [2] , a well-liked tree based classifier, is used to generate decision tree from a set of training examples.  ... 
doi:10.5121/ijist.2013.3103 fatcat:hhtzcojxgvavrndb3bif32oer4

Splitting matters: how monotone transformation of predictor variables may improve the predictions of decision tree models [article]

Tal Galili, Isaac Meilijson
2016 arXiv   pre-print
It is widely believed that the prediction accuracy of decision tree models is invariant under any strictly monotone transformation of the individual predictor variables.  ...  Accordingly, this study provides guidelines for both developers and users of decision tree models (including bagging and random forest).  ...  R package version 0.4-4. , 7 (2-3):81-227, 2012. Usama M Fayyad and Keki B Irani. On the handling of continuous-valued attributes in decision tree generation.  ... 
arXiv:1611.04561v1 fatcat:pehyxvda6ncgpkz7qsbnyqnfvm

Extracting Useful Rules Through Improved Decision Tree Induction Using Information Entropy

Mohd Mahmood Ali
2012 International Journal of Information Sciences and Techniques  
We suggest improvements to the existing C4.5 decision tree algorithm.  ...  Modified DMQL queries are used to understand and explore the shortcomings of the decision trees generated by C4.5 classifier for education dataset and the results are compared with the proposed approach  ...  C4.5 classifier [1] , [2] , a well-liked tree based classifier, is used to generate decision tree from a set of training examples.  ... 
doi:10.5121/ijist.2012.2608 fatcat:nseyoo6ybbdaplqbjsoswswvsu
« Previous Showing results 1 — 15 out of 3,464 results