Fuzzy Associative Classifier for Probabilistic Numerical Data
Recently, a number of advanced data collection and processing methodologies have led to the proliferation of uncertain data. When discovering from such uncertain data, we should handle these uncertainties with caution, because classical mining algorithms may not be appropriate for uncertain tasks. This paper proposes a generic framework of fuzzy associative classifier for probabilistic numerical data, which is prevalent in the real-world applications, such as sensor networks and GPS-based location. In this paper, we first introduce an Apriori-based algorithm for mining fuzzy association rules from a probabilistic numerical dataset based on novel support and confidence measures suitable for such dataset. Then, we give fuzzy rules redundancy pruning strategy and database coverage method to build a compact fuzzy associative classifier in removing redundant rules and thus improving the accuracy of the classifier. We also redefine multiple fuzzy rules classification method for classifying new instances. Extensive experimental results show the effectiveness and efficiency of our algorithm.
KeywordsFuzzy associative classifier Probabilistic numerical data Data mining
This research was supported by the National Basic Research Program of China (973 program) (2012CB316205), the National Natural Science Foundation of China (61070056, 61033010, 61202114), the HGJ Important National Science & Tech Specific Projects of China (2010ZX01042-001-002-002), and the Fundamental Research Funds of Renmin University of China (12XNLF07).
- 1.Qin XJ, Zhang Y, Li X, Wang Y (2010) Associative classifier for uncertain data. In: The 11th international conference on web-age information management (WAIM), pp 692–703Google Scholar
- 2.Qin B, Xia Y, Prbahakar S, Tu Y (2009) A Rule-based classification algorithm for uncertain data. In: IEEE international conference on data engineering (ICDE), pp 1633–1640Google Scholar
- 3.Qin B, Xia Y, Li F (2009) DTU: a decision tree for uncertain data. In: The Pacific-Asia conference on knowledge discovery and data mining (PAKDD), pp 4–15Google Scholar
- 5.Qin B, Xia Y, Li F (2010) A Bayesian classifier for uncertain data. In: ACM symposium on applied computing (SAC), pp 1010–1014Google Scholar
- 7.Gao CC, Wang JY (2010) Direct mining of discriminative patterns for classifying uncertain data. In: ACM SIGKDD international conference on knowledge discovery and data mining (KDD), pp 861–870Google Scholar
- 8.Liu B, Hsu W, Ma YM (1998) Integrating classification and association rule mining. In: ACM SIGKDD international conference on knowledge discovery and data mining (KDD) 1998, pp 80–86Google Scholar
- 9.Li WM, Han JW, Pei J (2001) CMAR: accurate and efficient classification based on multiple class-association rules. In: IEEE international conference on data mining, pp 369–376Google Scholar