A Comparative study of Stream Data mining Algorithms
Tusharkumar Trambadiya, Praveen Kumar Bhanodia · 2012
The problem to extract knowledge from large raw data has emerged as a new data structure. Data stream is a new era in data mining. Numerous algorithms are used for processing & classifying data streams. Traditional algorithms are not appropriate to process data stream which in cause generate problems regarding classification. A model which is developed from stream data for classification must update incrementally after the fresh arrival of new data. Data stream classification performance can be measured by various factors such as accuracy, computational speed, memory and time taken for processing. Data stream classification algorithm must have to meet certain requirements and measures to handle continuous flow of stream data. These algorithms get less time span to inspect data and build model, may be only once with less amount of resource, time and prediction. So to study the concepts of classification algorithms will lead towards development of better approaches for stream data mining. In this paper, we make a comparative study between Hoeffding tree, VFDT (Very Fast Decision Tree) and CVFDT (Concept Adaptinf Very Fast Decision Tree) from various algorithms which are used for stream data classification.