知识卡片
饱和激活只有局部近似线性
Sigmoid在有限输入上严格落于0和1之间,tanh落于−1和1之间,端点为极限。二者靠近原点可线性近似,却不能把−5到5定为精确线性区,或把其他范围说成唯一非线性区。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 深度学习模型
原书出处 · 1 处
- · 查看原书依据
Sigmoid: the sigmoid activation function is given by , where y is the input to the activation function and e ≈ 2.72 is the exponential constant. The domain of the sigmoid function is all real numbers (– ∞,+∞) and the range is bounded between 0 and 1 (0, +1). Specifically, for very small inputs (i.e., large negative values), the sigmoid is lower bounded by 0
参考来源
- 位置:《Machine Learning Fundamentals: Concepts, Models, and Applications》第08章《深度学习模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Machine Learning Fundamentals: Concepts, Models, and Applications/_epub-src/OEBPS/xhtml/21_chapter08.xhtml)
- 结论依据:摘录中关于饱和激活只有局部近似线性的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:Sigmoid: the sigmoid activation function is given by , where y is the input to the activation function and e ≈ 2.72 is the exponential constant. The domain of the sigmoid function is all real numbers (– ∞,+∞) and the range is bounded between 0 and 1 (0, +1). Specifically, for very small inputs (i.e., large negative values), the sigmoid is lower bounded by 0