A Data Mining Technique for Time Series Data
Yoshinori Sato, Akira Maeda · Industrial and Engineering Applications of Artificial Intelligence and Expert Systems · 2022
Data Mining is a technique to extract nontrivial and previously unknown regularities in databases. Data mining is one of the key technologies in realizing intelligent plant diagnosis system and quality control systems. In this paper, we propose a new data mining algorithm that uses time series data as input, and extracts meaningful relationships between trend patterns in time series data. In this algorithm, time series data is transformed to a set of trend pattern symbols. After that, the algorithm searches for interesting combinations of the symbols that may reveal a hidden causal relationship of the target system. Experimental results are also described to present the effectiveness of the proposed method.