A Moving-Window based Partial Periodic Patterns Update Technology in Time Series Databases

In the actual using, the data distribution of time series maybe changed with time. This dynamic behavior will led to the find pattern can´t be successful for the new data. Therefore, we present a partial periodic patterns update technology in time series databases based on the moving-window. The algorithm mines the patterns on the resent data in the moving-window, which only need to scan the data set in the moving-window two times mostly. The experiment results show that the new algorithm has more efficient than the nonmoving-window versions for the large databases.