本文深入探讨了神经网络的基本概念，包括神经元模型、逻辑单元、网络组件等关键部分。详细解析了多输出单位的One-vs-all策略，以及神经网络的学习过程，包括成本函数和反向传播算法。通过直观理解前向传播和反向传播算法，为读者提供了神经网络的全面理解。
吴恩达机器学习 Chapter9 Chapter10 Neural Network 神经网络
If there are too many features, then the hypothesis is non-linear. Neural network is an algorithm that mimics human’s brain.
Neuron model
Logistic unit
h
θ
(
x
)
=
1
1
+
e
−
θ
T
x
h_\theta(x)=\frac{1}{1+e^{-\theta^Tx}}
hθ(x)=1+e−θTx1
Sigmoid(logistic) activation function.
Neural Network
Component
- 3 layers:
- input layer
- hidden layer ≥ 1 \geq1 ≥1
- output layer
- a i ( j ) a_i^{(j)} ai(j) = “activation” of unit i in layer j
-
Θ
(
j
)
\Theta^{(j)}
Θ(j)=matrix of weights controlling function mapping from layer j to layer j+1
If network has s j s_j sj units in layer j j j, s j + 1 s_{j+1} sj+1 units in layer j + 1 j+1 j+1, then Θ ( j ) \Theta^{(j)} Θ(j) will be of dimension s j + 1 × ( s j + 1 ) s_{j+1}\times (s_j+1) sj+1×(sj+1).
Multiple output units: One-vs-all
Neural Network Learning
Cost function
Backpropagation Algorithm
Forward propagation
Vectorized computation
Backpropagation Algorithm
Algorithm
Intuition
δ
\delta
δ: a weighted sum of
δ
\delta
δ values weighted by corresponding edge strength.
扫一扫
2934
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
