摘要
运用现代多元统计技术——偏最小二乘回归PLS,在各影响因素存在多重共线性和样本数据短缺条件下进行了定量分析。本文不仅给出各影响因素对湖南电力消费的重要度指标图,而且确定这些影响因素与湖南电力消费需求之间的定量关系,为湖南电力市场分析提供了理论依据。
Applying modern statistical technology-Partial Least-Squares (PLS) regression, and quantity analyzing the relationship of Electricity power consumption on the condition of multi-collinearity between multivariable and small sample, the paper not only put forward the graph of variable importance projection about every factor relative to electricity power consumption, but also found the quantity relation of impact factors relative to power consumption, and set the frame of reference for electrical market analysis in Hunan Province.
出处
《系统工程》
CSCD
北大核心
2006年第9期98-102,共5页
Systems Engineering
关键词
偏最小二乘回归
多重共线性
影响因素分析
Partial Least Squares Regression
Multi-collinearity
Impact Factor Analysis