【摘要】 对从噪音信息中辨别有用的信息这一问题提出了一种解决方法。将大型的样本空间分解到可变的变量集和固定的变量集,并且让可变的变量集成为"工作集",使大型样本空间转变为小型的空间,然后引用primal-dual interior-point-solver来解决问题。更多还原

【Abstract】 A solution is proposed to tell the useful information from the noisy information.The large example space is decomposed into free variable set and fixed variable set,and the free variable set is made become the working set,which converts the large example space into a smaller one.Then,primal-dual interior-point-solver is applied to solve the problem.更多还原