Classification for sequential data involving human trial and error

This study proposes a novel classification method for sequential data invoving human trial and error. The classification of sequential data obtained from human experiments has become an important tool that supports the data analytics of the modern society. For example, several algorithms for motion recognition, voice recognition, etc. have been recently developed. The hidden Markov model (HMM) is a widely used probabilistic model that represents and analyzes such sequential data. To analyze the sequential data, it is necessary to consider the role of human-induced trial and error because it causes temporal and spatial variations in the observed trends. It is difficult to employ the conventional classification methods within the HMM for such sequential data. This is because the characteristics of the data change within the same class due to the variations. For such sequential data invoving spatio-temporal variations, it is effective to consider the class information and the characteristics of data together as a state in the HMM. Therefore, we propose a data classification method in this study by decoding the rich state to consider the impact of human trial and error.