查看“︁Bagging算法”︁的源代码

{{Expand language|1=en|time=2024-05-20T07:33:56+00:00}}
{{机器学习导航栏}}

'''Bagging算法''' （英語：'''B'''ootstrap '''agg'''regat'''ing'''，引導聚集算法），又稱'''裝袋算法'''，是[[機器學習]]領域的一種[[集成學習]][[算法]]。最初由[[Leo Breiman]]於1994年提出。Bagging算法可與其他[[統計分類|分類]]、[[回歸分析|回歸]]算法結合，提高其準確率、穩定性的同時，透過降低結果的[[變異數]]，避免[[過擬合]]的發生。

==算法步骤==
给定一个大小为<math>n</math>的[[训练集]]<math>D</math>，Bagging算法从中均匀、有放回地（即使用[[自助法|自助抽样法]]）选出<math>m</math>个大小为<math>n'</math>的[[子集]]<math>D_i</math>，作为新的训练集。在这<math>m</math>个训练集上使用分类、回归等算法，则可得到<math>m</math>个模型，再透過取[[平均值]]、取多数票等方法，即可得到Bagging的结果。

==参考文献==
* {{Cite journal 
 | last = Breiman 
 | first = Leo 
 | authorlink = Leo Breiman 
 | title = Bagging predictors 
 | journal = [[Technical Report No. 421]] 
 | year = 1994 
 | url = https://www.stat.berkeley.edu/~breiman/bagging.pdf 
 | access-date = 2020-01-11 
 | archive-date = 2019-07-18 
 | archive-url = https://web.archive.org/web/20190718072616/https://www.stat.berkeley.edu/~breiman/bagging.pdf 
 | dead-url = no 
 }}
* {{Cite journal
 | last = Breiman 
 | first = Leo 
 | authorlink = Leo Breiman 
 | title = Bagging predictors 
 | url = https://archive.org/details/sim_machine-learning_1996-08_24_2/page/123
 | journal = [[Machine Learning (journal)|Machine Learning]] 
 | volume = 24 
 | issue = 2 
 | pages = 123–140 
 | year = 1996 
 | id = {{citeseerx|10.1.1.32.9399}}
 | doi = 10.1007/BF00058655 
}}
* {{Cite journal 
 | last = Alfaro 
 | first = E., Gámez, M. and García, N. 
 | title = adabag: An R package for classification with AdaBoost.M1, AdaBoost-SAMME and Bagging 
 | year = 2012 
 | url = https://cran.r-project.org/package=adabag 
 | journal =  
 | access-date = 2016-10-02 
 | archive-date = 2019-09-24 
 | archive-url = https://web.archive.org/web/20190924162440/https://cran.r-project.org/web/packages/adabag/index.html 
 | dead-url = no 
 }}

[[Category:集成学习]]
[[Category:機器學習演算法]]
[[Category:計算統計學]]