BoosTexter: A Boosting-based Systemfor Text Categorization
Machine Learning - Special issue on information retrieval
Knowledge Discovery in Multi-label Phenotype Data
PKDD '01 Proceedings of the 5th European Conference on Principles of Data Mining and Knowledge Discovery
Data Mining: Concepts and Techniques
Data Mining: Concepts and Techniques
Multilabel Neural Networks with Applications to Functional Genomics and Text Categorization
IEEE Transactions on Knowledge and Data Engineering
ML-KNN: A lazy learning approach to multi-label learning
Pattern Recognition
Statistical Comparisons of Classifiers over Multiple Data Sets
The Journal of Machine Learning Research
Random k-Labelsets: An Ensemble Method for Multilabel Classification
ECML '07 Proceedings of the 18th European conference on Machine Learning
Multi-label Classification with Gene Expression Programming
HAIS '09 Proceedings of the 4th International Conference on Hybrid Artificial Intelligence Systems
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
The WEKA data mining software: an update
ACM SIGKDD Explorations Newsletter
Random block coordinate descent method for multi-label support vector machine with a zero label
Expert Systems with Applications: An International Journal
| Hi-index | 12.05 |
Some of the real-world problems are represented with just one label but many of today's issues are currently being defined with multiple labels. This second group is important because multi-label classes provide a more global picture of the problem. From the study of the characteristics of the most influential systems in this area, MlKnn and RAkEL, we can observe that the main drawback of these specific systems is the time required. Therefore, the aim of the current paper is to develop a more efficient system in terms of computation without incurring accuracy loss. To meet this objective we propose MlCBR, a system for multi-label classification based on Case-Based Reasoning. The results obtained highlight the strong performance of our algorithm in comparison with previous benchmark methods in terms of accuracy rates and computational time reduction.