Online learning of electric vehicle consumers' charging behavior with missing data

Learning in the presence of missing data is a pervasive problem in statistical data analysis. This paper deals with tracking the dynamic charging behavior of electric vehicle consumers, when some of the consumers' consumption decisions are missing. The problem is then formulated as an online classification task with missing labels. An online algorithm is proposed to jointly impute the missing data while at the same time learn from the complete data using an online convex optimization approach.