Improvised volatility shift detector by drift detection in data streams

Drift detection techniques are new concept which detects a change in distribution within a stream. In the streaming environment, it is the process of segmenting a data stream into different segments. These segments are obtained by identifying the points where the stream dynamically changes. At present there are no such techniques that analyze the change in the rate of these detected changes. The term “stream volatility” was proposed to describe the rate of changes in a stream. Efficient method was proposed to find volatility of change points detected by drift detector. Variance in change point verifies volatility in data streams. The random sampling method is used to calculate volatility but downside of random sampling method used for finding volatility is that some points are missed out while calculating volatility. These missing points may cause huge difference in volatility measure. To overcome this problem, we propose window technique which would consider all points concern to measure volatility without losing any more recent point.