CAJ | 학술논문

当变量间存在多重相关性时，常采用偏最小二乘（partial least squares，PLS）回归进行建模。但是，传统 PLS 回归采用线性关系式来建立内部成分与外部成分之间的关系，已有的非线性 PLS 在模型可解释性方面存在不足。针对军用软件成本估算问题的非线性特点，建立了一种基于内部机理的非线性 PLS 回归模型，给出了模型的推导过程和实现算法。实例分析表明，该方法的估算精度优于多元线性回归，线性 PLS 回归和基于多项式内部映射的非线性回归。
당변량간존재다중상관성시，상채용편최소이승（partial least squares，PLS）회귀진행건모。단시，전통 PLS 회귀채용선성관계식래건립내부성분여외부성분지간적관계，이유적비선성 PLS 재모형가해석성방면존재불족。침대군용연건성본고산문제적비선성특점，건립료일충기우내부궤리적비선성 PLS 회귀모형，급출료모형적추도과정화실현산법。실례분석표명，해방법적고산정도우우다원선성회귀，선성 PLS 회귀화기우다항식내부영사적비선성회귀。
For the multi-correlation of variables,partial least squares (PLS)regression is often adopted for modeling.However,traditional PLS establishes a linear relationship between inner and outer latent compo-nents,and existing nonlinear PLS models have some weakness in interpretability.Aiming at the nonlinearity in military software cost estimation,a nonlinear PLS regression model based on inner mechanism is established,and the derivation process and the arithmetic are presented.The experimental results show that the estimation accuracy is supe-rior to multiple linear regression,linear PLS regression and polynomial nonlinear PLS regression.