Machine Learning Series No.6 -- EM algorithm
EM算法 1.直觀理解 通俗理解：http://www.javashuo.com/article/p-asqtxiud-mh.html 通俗的理解看出就是EM算法由於不知道隱變量的分佈，先給出參數的隨機初始值，然後根據參數，去得到隱變量的分佈，然後根據隱變量和觀測變量的共同分佈基於最大似然去重新估計參數，知道參數穩定。 2.數學推導 極大似然估計： L(θ)=∑ilogp(xi;θ)=∑ilog(
相關文章
- 1. Machine Learning Series No.4 -- Decision Trees
- 2. Machine Learning Series No.10 -- PCA
- 3. Machine Learning Series No.7 -- Max Entropy
- 4. Machine Learning Series No.2 --Logistic Regression
- 5. Machine Learning Series No.3-- Support Vector Machine
- 6. machine learning[GMM-EM]
- 7. Machine Learning Series No.8 -- CRF(Conditional Random field)
- 8. Machine Learning in Action:KNN Algorithm
- 9. Machine Learning(8): Reinforcement learning algorithm
- 10. Machine learning(6): Quiz 6
- 更多相關文章...
- • Docker Machine - Docker教程
- • Redis的6種數據類型 - Redis教程
- • Java Agent入門實戰（一）-Instrumentation介紹與使用
- • Java Agent入門實戰（三）-JVM Attach原理與使用
相關標籤/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。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. Machine Learning Series No.4 -- Decision Trees
- 2. Machine Learning Series No.10 -- PCA
- 3. Machine Learning Series No.7 -- Max Entropy
- 4. Machine Learning Series No.2 --Logistic Regression
- 5. Machine Learning Series No.3-- Support Vector Machine
- 6. machine learning[GMM-EM]
- 7. Machine Learning Series No.8 -- CRF(Conditional Random field)
- 8. Machine Learning in Action:KNN Algorithm
- 9. Machine Learning(8): Reinforcement learning algorithm
- 10. Machine learning(6): Quiz 6