Evolutionary semi-supervised learning with swarm intelligence

To address the issue of evolutionary data classification, we propose an evolving swarm classification model. It treats each class as an ant colony carrying different type of pheromone. The ant colonies send their members to propagate their unique pheromone on the unlabeled instances, so as to label them for member recruitment. Meanwhile, the unlabeled instances are treated as unlabeled ants, which also have their preferences for joining one of those labeled colonies. We call it homing feedback, and integrate it into the pheromone update process. Afterwards, the natural selection process is carried out to keep a balance between the member recruitment and the ant colony size maintenance. Sufficient experiments demonstrate that our algorithm is effective in the real-world evolutionary classification applications.

[1]  Latifur Khan,et al.  Facing the reality of data stream classification: coping with scarcity of labeled data , 2012, Knowledge and Information Systems.

[2]  Li Guo,et al.  Classifier and Cluster Ensembles for Mining Concept Drifting Data Streams , 2010, 2010 IEEE International Conference on Data Mining.

[3]  Christoforos Anagnostopoulos,et al.  Temporally adaptive estimation of logistic classifiers on data streams , 2009, Adv. Data Anal. Classif..

[4]  Ludmila I. Kuncheva,et al.  On the window size for classification in changing environments , 2009, Intell. Data Anal..

[5]  Li Guo,et al.  A framework for application-driven classification of data streams , 2012, Neurocomputing.

[6]  Shuicheng Yan,et al.  Semi-Supervised Classification on Evolutionary Data , 2009, IJCAI.

[7]  Philip S. Yu,et al.  Classifying Data Streams with Skewed Class Distributions and Concept Drifts , 2008, IEEE Internet Computing.

[8]  Concha Bielza,et al.  Classifying evolving data streams with partially labeled data , 2011, Intell. Data Anal..

[9]  Li Guo,et al.  Enabling Fast Lazy Learning for Data Streams , 2011, 2011 IEEE 11th International Conference on Data Mining.

[10]  Li Guo,et al.  Enabling fast prediction for ensemble models on data streams , 2011, KDD.

[11]  Qinghua Hu,et al.  Dynamic classifier ensemble using classification confidence , 2013, Neurocomputing.

[12]  Jason Lines,et al.  A shapelet transform for time series classification , 2012, KDD.

[13]  Yun Chi,et al.  Evolutionary spectral clustering by incorporating temporal smoothness , 2007, KDD '07.

[14]  Haixun Wang,et al.  A Low-Granularity Classifier for Data Streams with Concept Drifts and Biased Class Distribution , 2007, IEEE Transactions on Knowledge and Data Engineering.

[15]  Deepayan Chakrabarti,et al.  Evolutionary clustering , 2006, KDD '06.