Data Quality Identification Model for Power Big Data

Data quality identification is an important task in power big data.Abnormal data exist and hamper the effective utilization of power big data.Moreover,the lack of labeled data makes the detection of abnormal data more challenging.Then,a data quality identification model for power big data is proposed.It can detect abnormal data from massive power big data.In this model,power data are grouped and then mapped into different feature spaces based on data augmentation technology.Tri-training is applied to detect abnormal data from different power data from different feature spaces.Experiments and simulations are performed to demonstrate the effectiveness of the proposed model.