A REVIEW ON PREDICTION OF MISSING DATA IN MULTIVARIABLE TIME SERIES
P Vikas Kumar, Mukkamula Venu Gopalachari · INTERNATIONAL JOURNAL OF COMPUTER APPLICATION · 2017
Time series models helps to track the behaviour for process or metrics over a set of time periods.The advancements in the complexity of the time series data leads to the evolution of multi variable time series analysis demanding qualitative machine learning techniques.Data missing in collections of multi variable time series occurs frequently is a challenging issue, due to the involvement of sensors to generate huge amount of data, which affects the quality of the data analysis.In order to overcome this issue, there is a need of applying efficient missing data prediction models.This paper addresses the impact of missing data in multi variable time series and presents the survey of the various models exists to handle missing data.The focus of this paper is on matrix factorization models in smoothing time series data and also the future directions to improve the quality of the missing data prediction.