中文摘要：

以非可加模糊测度代替经典可加测度,基于模糊积分建立非线性回归模型是新近出现的数据建模方法。该方法充分考虑自变量因素之间的信息熔合（含协同或冲突）作用。本文完整地给出了适用于实数范围内的基于模糊积分（含Choquet积分和∨Sipo∨s积分）的多元非线性回归模型转化为普通线性回归模型的非线性转换方法及其简化算法。并将该方法应用于金融市场数据分析,结果表明效果较之普通多元线性回归有大的提高,且方法简便容易应用。

英文摘要：

There is new nonlinear multi-regressions method for data modeling based on fuzzy integral,in which additive measure is substituted by non-additive measure,which consider the information fusion among independent variables,including synergy and conflict.In the paper,the full model of nonlinear multi-regressions based on fuzzy integral（both Choquet integral ipo integral included） is presented, and one nonlinear transformation for transforming the nonlinear model into linear one under real number is given.And the reduced arithmetic for nonlinear transformation is presented,which is applicable.Then this model is applied to financial data analysis,which turn out that it is improved comparative to linear multi-regressions.