已有的一些判别训练(discriminative training)方法如Boosting为了提高算法的效率,要求损失函数(loss function)是可以求导的,这样的损失函数无法体现最直接的优化目标.而根...
基于20个网页-相关网页
在特征向量的选择中,本算法还用到了训练集的类别标签和类别平均向量的判别信息。
In selecting vectors, the algorithm also USES the class label of training set and the judgment information of class mean vector.
设计了一种英文字符联机识别的典型方法,其基本原理为大量训练数据的统计判别。
A typical algorithm of English character machine -linked recognition, based on statistical distinction of enormous drill data.
倘若条件独立性假设确实满足,朴素贝叶斯分类器将会比判别模型,譬如逻辑回归收敛得更快,因此你只需要更少的训练数据。
If the NB conditional independence assumption actually holds, a Naive Bayes classifier will converge quicker than discriminative models like logistic regression, so you need less training data.
应用推荐