WSEAS Transactions on Computers
Print ISSN: 1109-2750, E-ISSN: 2224-2872
Volume 15, 2016
Mining Data Streams with Concept Drift in Massive Online Analysis Frame Work
Authors: ,
Abstract: The advancement of the technology has resulted in the data generation with increasing rate of data distribution. The generated data is called as ’data stream’. Data streams can be mined only by using sophisticated techniques. The stream data mainly comes from mobile applications, sensor applications, network monitoring, traffic management, weblogs etc. But the concepts often change with time. Weather forecasting data is a good examples here. The model built on old data is inconsistent with the new data and regular updation of the model is necessary. This type of change in a data stream is called as concept drift. The paper aims at mining data streams with concept drift in Massive Online Analysis Frame work by using Naive Bayes algorithm using classification technique. The authors also generated their own data set generator OURGENERATOR for the analysis. The other generators used are LED, RANDOMRBF, WAVEFORM, SEA, STAGGER and HYPERPLANE with concept drift. Along with Our Generator the other three static generators used are: Electricity, Airline and Forest Cover. The performance of the Naive Bayes on RANDOMRBF generator is found to be excellent but equally it is best on OUR_GENERATOR also which is first of its kind in the literature.
Search Articles
Pages: 133-142
WSEAS Transactions on Computers, ISSN / E-ISSN: 1109-2750 / 2224-2872, Volume 15, 2016, Art. #14