Spatial Air Quality Prediction Using Gaussian Process

Abstract

The exponential development of economy has evoked the problem of environment, particularly the PM2.5, which is extremely detrimental to people’s health and has gained high attention from the public. Under contemporary circumstance, current software that report the level of PM2.5 is limited by the city level. Thus, with the intension to help people aware the exact air quality of surrounding areas, we installed 280 fine designed devices to these places and presented a Gaussian Process based inference model to estimate the value at any place. Based on the real data and compared to related methods, the experimental results of proposed method prove the effectiveness of it.