The Community for Technology Leaders
RSS Icon
Subscribe
Honolulu, HI, USA USA
June 24, 2012 to June 29, 2012
ISBN: 978-1-4673-2892-0
pp: 694-701
ABSTRACT
We propose a novel class-based micro-classifier ensemble classification technique (MCE) for  classifying data streams. Traditional  ensemble-based data stream classification techniques build  a classification model from each data chunk and  keep an ensemble of such models. Due to the fixed length of  the ensemble, when a new model is trained, one existing model is discarded. This creates several problems. First, if a  class disappears from the stream and reappears after a long  time, it would be misclassified if a majority of the classifiers  in the ensemble does not contain any model of that class. Second,  discarding a model means discarding the corresponding data  chunk completely. However, knowledge obtained from some classes  might be still useful and if they are discarded, the overall error rate would increase. To address these problems, we propose an ensemble model where each class information is stored separately.  From each data chunk, we train a model for each class of data.  We call each such model a micro-classifier. This approach is more robust than existing chunk-based  ensembles in handling dynamic changes in the data stream.  To the best of our knowledge, this is the first attempt to  classify data streams using the class-based ensembles approach.  When the number of classes grow in the stream, class-based ensembles may degrade in performance (speed). Hence, we sketch a cloud-based solution of our class-based ensembles to handle a large number of classes effectively. We compare our technique with several state-of-the-art data  stream classification techniques on both synthetic and benchmark data streams, and obtain much higher accuracy.
INDEX TERMS
Data models, Training, Testing, Training data, Accuracy, Data mining, Cloud computing, classification MapReduce Ensemble cloud
CITATION
Tahseen M. Al-Khateeb, Mohammad M. Masud, Latifur Khan, Bhavani Thuraisingham, "Cloud Guided Stream Classification Using Class-Based Ensemble", CLOUD, 2012, 2013 IEEE Sixth International Conference on Cloud Computing, 2013 IEEE Sixth International Conference on Cloud Computing 2012, pp. 694-701, doi:10.1109/CLOUD.2012.127
43 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool