Journals Proceedings

Abstract

Towards the end of the 20th century, we have seen an improved interest among Statisticians and Computer engineers to explore data from any data source with respect to the change in time. However, most of the techniques used remains the same as that used in conventional data mining. Capturing, indexing, representing and storing the data remains the key issue in time series data mining. Indexing is a very critical under job under noise conditions. The indexing system exploded the database volume. In time series data ming a statistical models which provides descriptions for the sampling of data, (data collected on global warming, flood and flood forecasting pattern etc) are deviced. In order to provide a statistical arrangement for describing the nature of a continues stream of data that fluctuate in a random fashion with respect to the time, we assume a time series can be defined as a collection of random variables indexed according to the order they are obtained. Here we are assuming a time series data as a sequence of (linear or non-linear) random variables, which can be represented as t1,t2,t3,t4 …., where the random variables t1,t2 etc. are the observed values with respect to the change in time. Trend detection and recording is the most important activity in time series data mining. In practice, it is accomplished using linear and nonlinear regression technique that satisfactorily helps in identifying non-monotonous trend component in the time series. It is already been proved that statistical methods such as moving average can be effectively used in smoothening data flow.