Outlier detection of air temperature series data using probabilistic finite state automata‐based algorithm
Jun Shen, Minhua Yang, Bin Zou, Neng Wan, Yu-Fang Liao · Complexity · 2012
Abstract This article proposes a probability finite state automata‐based algorithm (PFSAA) for detecting outliers of air temperature series data caused by sensor errors. This algorithm first divides the training samples of air temperature series data into subclusters that will be further used to build finite state automata by splitting and combining techniques. Then, it creates a dynamic transition matrix of PFSA based on probability theories. Finally, the outliers of the remaining test samples are detected by PFSAA. The proposed algorithm is quantitatively validated by the reference data and a traditional backpropagation neural net model. © 2012 Wiley Periodicals, Inc. Complexity, 2012