Machine Learning Techniques 笔记：2-8 Ada Boosting
先算出错误率Epselo t，把犯错误的点做（1-Et）的放缩，正确的点做Et比例的放缩=》正确的乘以♦t，错误的除以♦t 通常，♦t≥1，因为1/2是丢铜板乱猜的概率。只要base algroithm比乱猜好，♦t大于等于1，使得正确的得以缩小，错误的得以放大，更加专注在犯错误的点上面，来得到不一样的hypothesis。有点类似于PLA，有错即改 只要比乱猜效果好，最终Ein总可收敛 t=1，
相关文章
- 1. Machine Learning Techniques 笔记：2-13 Deep Learning
- 2. Machine Learning Techniques 笔记：2-6 Support Vector Regression
- 3. Machine Learning Techniques 笔记：2-12 Neural Network
- 4. Machine Learning Techniques 笔记：2-15 Matrix Factorization
- 5. Machine Learning Techniques 笔记：2-5 Kernel Logistics Regression
- 6. Machine Learning Techniques 笔记：2-7 Blending and Bagging
- 7. Machine Learning Techniques 笔记：2-9 Decision Tree
- 8. Machine Learning Techniques 笔记：2-10 Random Forest
- 9. Machine Learning Techniques 笔记：2-2 Dual Support Vector Machine
- 10. Machine Learning Techniques 笔记：2-3 Kernel Support Vector Machine
- 更多相关文章...
- • Docker Machine - Docker教程
- • ASP.NET Razor - 标记 - ASP.NET 教程
- • Tomcat学习笔记（史上最全tomcat学习笔记）
- • Java Agent入门实战（一）-Instrumentation介绍与使用
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. Machine Learning Techniques 笔记：2-13 Deep Learning
- 2. Machine Learning Techniques 笔记：2-6 Support Vector Regression
- 3. Machine Learning Techniques 笔记：2-12 Neural Network
- 4. Machine Learning Techniques 笔记：2-15 Matrix Factorization
- 5. Machine Learning Techniques 笔记：2-5 Kernel Logistics Regression
- 6. Machine Learning Techniques 笔记：2-7 Blending and Bagging
- 7. Machine Learning Techniques 笔记：2-9 Decision Tree
- 8. Machine Learning Techniques 笔记：2-10 Random Forest
- 9. Machine Learning Techniques 笔记：2-2 Dual Support Vector Machine
- 10. Machine Learning Techniques 笔记：2-3 Kernel Support Vector Machine