An Improved Twin Support Vector Regression Machine Based on Adjusted Cosine Similarity for Load Prediction

Autor: Yufeng Wu, Chen Xiao, Dong Liu, Yun Rao, Song Yao, Qingyang Liao
Rok vydání: 2021
Předmět:
Zdroj: 2021 China International Conference on Electricity Distribution (CICED).
DOI: 10.1109/ciced50259.2021.9556774
Popis: Due to the rapid increase of power data resources on the consumer side, using of large-scale data to achieve fast and high-precision load prediction has become an urgent problem. This paper proposes an improved twin support vector regression machine for load forecasting based on adjusted cosine similarity. The data sparsity of the improved twin support vector regression machine is better than the twin support vector regression machine which is the novel machine learning method. The training speed of the improved twin support vector regression machine is faster than the support vector regression machine which is the classic machine learning method. After verification of the actual grid data, this method has good prediction accuracy and prediction speed compared with the twin support vector regression machine and the support vector regression machine.
Databáze: OpenAIRE