Efficient temporal pattern mining in big time series using mutual information

Very large time series are increasingly available from an ever wider range of IoT-enabled sensors deployed in different environments. Significant insights can be gained by mining temporal patterns these series. Unlike traditional pattern mining, (TPM) adds event intervals into extracted patterns, making them more expressive at the expense increased and space complexities. Existing TPM methods either cannot scale to datasets, or work only on pre-processed events rather than This paper presents our Frequent Temporal Pattern Mining Time Series (FTPMfTS) approach providing: (1) The end-to-end FTPMfTS process taking as input producing frequent output. (2) efficient Hierarchical Graph (HTPGM) algorithm that uses data structures for fast support confidence computation, employs effective pruning techniques significantly faster mining. (3) An approximate version HTPGM mutual information, a measure correlation, prune unpromising search space. (4) extensive experimental evaluation showing outperforms baselines runtime memory consumption, big datasets. is up two orders magnitude less consuming baselines, while retaining high accuracy.