An APCA-enhanced compression method on large-scale time-series data

With the widespread use of sensors, large-scale time series data becomes ubiquitous. As a result, it is important to provide efficient compression storage and retrieval algorithms, which are the base of subsequent data mining analysis, on these data. One of the most widely used data compression algorithms on time series data is Adaptive Piecewise Constant Approximation (APCA). However, APCA's compression storage overhead is still great for large-scale time series data. In this paper, we present a novel APCA-Enhanced algorithm. Extensive experiments over large real-world data sets demonstrate our algorithm's better performance of compression ratio and query latency compared with run length encoding and APCA.