The prediction of user attributes based on smart grid big data is of great significance to the construction of smart grid analysis systems and intelligent building construction. Traditional machine learning methods for single-user attribute analysis cannot use the relationship between attributes to improve the accuracy, and can not easily mine the missing data. These two problems have restricted the design of smart grid systems and the improvement of smart building systems.
Cong Yang, a researcher at the Shenyang Institute of Automation, Chinese Academy of Sciences, based on the study of multi-year machine learning algorithms, proposed a supervised/semi-supervised user attribute prediction model based on multi-task learning to achieve a small amount of available data by treating each attribute prediction as a single task. Multiple tasks simultaneously learn and make decisions. At the same time, the relationship between multiple user attributes was excavated, the accuracy of multiple attribute predictions was improved, and the missing data sample information was fully used to further improve the model generalization ability.
Relevant research results were published on the IEEE Transactions on Smart Grid and Pattern Recogniton, respectively, using the Joint Household Functional Prediction via Smart Meter Data and User attribute discovery with missing labels. The research work was supported by the State Key Laboratory of Robotics and the National Natural Science Foundation of China.
Lacquer Kitchen Cabinet,Lacquer Cabinets,Lacquer Finish Cabinets,High Gloss Lacquer Kitchen Cabinets
Ningbo Oulin Import&Export Co.,Ltd. , https://www.oulinfurniture.com