知识卡片
循环状态传递过去信息
逐帧无状态的局部网络只能使用其感受野内的信息;循环网络把先前处理结果编码到状态,再用于后续输入。状态可使昏暗帧参考早先清晰观测,但能否保存相关内容取决于容量和训练。前向网络若显式输入历史也能利用过去,区别在于信息如何被提供和更新。
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
- · 循环神经网络
原书出处 · 1 处
- · 查看原书依据
RNNs are all about adaptation. They solve the following problem: What if we want our computations in the past to influence our computations in the future? If we have a stream of data, a feedforward net has no way to adapting its behavior based on what outputs it previously produced.
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第25章《循环神经网络》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c025.xhtml)
- 结论依据:摘录中关于循环状态传递过去信息的定义、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:RNNs are all about adaptation. They solve the following problem: What if we want our computations in the past to influence our computations in the future? If we have a stream of data, a feedforward net has no way to adapting its behavior based on what outputs it previously produced.