Find centralized, trusted content and collaborate around the technologies you use most.
Teams
Q&A for work
Connect and share knowledge within a single location that is structured and easy to search.
是否可以向训练集中添加一个新的数据点,使 SVM 损失保持不变,但 Softmax 分类器损失不变?
在 SVM 中,如果新数据点的分数超出正确类别分数的边际范围,则损失不会改变,但在 Softmax 损失中,如果新添加的数据点的分数接近 +infinity,则会产生不利影响影响损失,但绝对Softmax的损失会改变。