Revision-Linear Regression
Classfication
The MNIST dataset
train: 训练集还是测试集
The CIFAR-10 dataset
1. Regression VS Classfication
输出概率
1.1 How to map [0,1]
导数: 正态分布
1.2 Sigmoid functions
2. Logistic Regression model
loss function for Binary Classification
Cross-entropy 交叉熵
两个分布中间差异性大小 越大越好
BCE
Mini-Batch loss function for binary classification
分布接近
3. Implementation
Total:
result: