A multi-label classification method based on associative rules

作者：Zhu Yuquan^*; Luo Wei; Chen Geng; Ou Jishun

来源：Journal of Computational Information Systems, 2012, 8(2): 791-799.

• Associative rule
• Audio classification
• Classification methods
• Corresponding solutions
• Frequent itemset
• Gene-function analysis
• Higher efficiency
• Mining frequent itemsets
• Multi-label
• Recognition rates
• Research directions
• Sample Database
• Training sample

In the practical application of image classification, text and audio classification and gene function analysis, the data that classification faces has presented multi-label feature, multi-label classification has become an important research direction of classification. Based on this, we put forward a multi-label classification method named MLCMBAR which is based on associative rules, this method constructs the accurate multi-label classification associative rules by mining frequent itemsets in the sample database and corresponding solutions of some key problems that appear in the process of mining are put forward. Compared to the existing multi-label classification associative rules mining algorithms used in training sample database, The experimental results show that MLCMBAR has higher efficiency and recognition rate.

• 出版日期2012