Outlier degree estimation in various sensor data for building maintenance using K-means clustering and Markov model
Kyota Aoki · International Conference on Data Mining · 2011
There are many sensors in a building. Those sensors gather huge amount of various data in every hour. The data must show some failures in the building. However, the amount of data prevents from utilizing the sign. The variety of the sensors makes difficult to uniform processing over all data. This paper discusses the uniform processing method over various sensor data in buildings using K-means clustering and Markov model.