知识卡片

普通读书笔记卡 · 9170 · 生成模型:GPT-6

非线性隐藏层扩大可分结构

理解等级:中级

XOR四点无法由一个线性超平面分开,多个隐藏单元先形成中间划分,再由输出组合可表达它。仅堆叠线性层仍等效于单个仿射映射,增加层数本身不足以获得非线性决策边界。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      We already know that a single perceptron can create a single separating hyperplane; we will therefore need at least two perceptrons to solve the XOR problem. Using two perceptrons in parallel will yield tuples of the kind (0,0), (0,1), and so on; hence, we then need another perceptron to recombine these tuples into a single output. Figure 10.5 shows the simplest multilayer perceptron that can

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第10章《人工神经网络》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/chapter_10.xhtml) - 结论依据:摘录中关于非线性隐藏层扩大可分结构的定义、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:We already know that a single perceptron can create a single separating hyperplane; we will therefore need at least two perceptrons to solve the XOR problem. Using two perceptrons in parallel will yield tuples of the kind (0,0), (0,1), and so on; hence, we then need another perceptron to recombine these tuples into a single output. Figure 10.5 shows the simplest multilayer perceptron that can