We introduce a new model of membership query (MQ) learning, where the
learning algorithm is restricted to query points that are \emph{close} to
random examples drawn from the underlying distribution. The learning model is
intermediate between the pac model (Valiant, 1984) and the PAC+M
本文提出了一种学习带有随机分类噪声下奇偶函数的略微次指数时间算法,并给出了有关在 PAC(Probably Approximately Correct)模型中证明不可学习概念类别的噪声容错算法。同时,对于 Kearns 的统计查询模型可证明不可学习的概念类,我们给出了一种有效的噪声容错算法。此外,文章还探讨了基于统计查询模型扩展到 t 元组查询的情况,从而证明该扩展不会增加可弱学习函数的集合。