不确定性序回归模型在普通高校扩招预警中的应用

A New Uncertainty Ordinal Regression Method about Early Warning of College Enrollment

根据专家决策和结构风险最小化原则,构建了一种新的不确定性有序支持向量回归模型,可以解决训练点带有不确定性的序回归问题。基于OSVR首先构建了一个较为复杂的优化模型;该模型的约束条件比较多,特别是在海量数据的情况下处理起来就更为麻烦,提出一个等价优化问题,并进行了由自然数到实数的推广,将复杂的模型转换为一个相对简单的优化模型;然后利用技巧将线性学习问题很自然地拓广到非线性情况,实现了从训练样本空间到高维特征空间的映射。普通高校招生人数预警的数据试验表明模型有一定的实际应用价值。

A new machine learning technique is proposed,which is able to deal with training data with uncertainty based on expert advices.The problem of predicting variables of ordinal scale is referred to as ordinal regression and is complementary to the standard machine learning tasks of classification and metric regression.However,it is required that every input must be exactly assigned to one of these classes without any uncertainty.Firstly,a complicated model is proposed based on ordinal regression.But it is difficult in solving the problem with many restrictions especially for lager data.And an equal function is produced which is simpler.Then by kernel game,nonlinear model is introduced.Moreover,the problem about early warning of college enrollment is solved by our algorithm.