Logistic Regression(逻辑斯蒂回归)Classification(分类问题)使用1和0,或者positive和negative表示事件的相对立的两种可能情况。(这里指二元分类问题) e.g. 垃圾邮件与非垃圾邮件等 Logistic Regression Model(逻辑斯蒂回归模型)Sigmoid Function(又Logistic Function)
Hypothesis Output(预测函数)
Decision boundary(决策边界)Cost Function(损失函数)
Simplified Cost Function(简化损失函数)参数优化Gredient Decent(梯度离散下降)参数更新与表达Advanced OptimizationMultu-class Classfication:One vs AllRegularization(正则化)Overfitting & Underfitting(过度拟合与欠拟合)相关概念
e.g. Adressing Overfitting(过度拟合的解决方法)
RegularizationCost Function(含正则化的损失函数)Regularized Linear Reguression(正则化的线性回归)
Regularized Logistic Reguression(正则化的逻辑斯蒂回归)
来源:http://www./content-4-155451.html |
|