Fuzzy inductive reasoning forecasting strategies able to cope with missing data: A smart grid application

详细信息    查看全文

• 作者：Sergio Jurado^a ; ^{sergio.jurado@sensingcontrol.com} ; ^{s.juradogomez@gmail.com} ; À ; ngela Nebot^b ; ^{angela@cs.upc.edu} ; Fransisco Mugica^b ; ^{fmugica@cs.upc.edu} ; Mihail Mihaylov^c ; ^{mmihaylo@vub.ac.be}
• 关键词：Soft computing ; Fuzzy inductive reasoning ; Entropy-based feature selection ; Prediction with missing values ; Energy modelling
• 刊名：Applied Soft Computing
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：51
• 期：Complete
• 页码：225-238
• 全文大小：1773 K
• 卷排序：51

文摘

We developed an improved version of FIR(Fuzzy Inductive Reasoning) that can cope with MVs(v): flexible FIR. Eight different strategies of prediction for flexible FIR are proposed. We compare with energy real data from eight buildings with different profiles. Better results are obtained in those strategies making use of causal relevance. Flexible FIR predicts 97% of registers when 73% of training data have MVs.