1、原理

https://www.cnblogs.com/zhouxiaohui888/p/6008368.html

2、实战

xgboost中比较重要的参数介绍:

(1)学习率:learning rate :一般设置比较低,0.1以下

(2)tree:

max_depth

min_child_weight

subsample

colsample_bytree

gamma

(3)正则化参数

lambda

alpha

(1)objective [ default=reg:linear ] 定义学习任务及相应的学习目标,可选的目标函数如下:

  • “reg:linear” –线性回归。
  • “reg:logistic” –逻辑回归。
  • “binary:logistic” –二分类的逻辑回归问题,输出为概率。
  • “binary:logitraw” –二分类的逻辑回归问题,输出的结果为wTx。
  • “count:poisson” –计数问题的poisson回归,输出结果为poisson分布。 在poisson回归中,max_delta_step的缺省值为0.7。(used to safeguard optimization)
  • “multi:softmax” –让XGBoost采用softmax目标函数处理多分类问题,同时需要设置参数num_class(类别个数)
  • “multi:softprob” –和softmax一样,但是输出的是ndata * nclass的向量,可以将该向量reshape成ndata行nclass列的矩阵。没行数据表示样本所属于每个类别的概率。
  • “rank:pairwise” –set XGBoost to do ranking task by minimizing the pairwise loss

(2)’eval_metric’ The choices are listed below,评估指标:

  • “rmse”: root mean square error
  • “logloss”: negative log-likelihood
  • “error”: Binary classification error rate. It is calculated as #(wrong cases)/#(all cases). For the predictions, the evaluation will regard the instances with prediction value larger than 0.5 as positive instances, and the others as negative instances.
  • “merror”: Multiclass classification error rate. It is calculated as #(wrong cases)/#(all cases).
  • “mlogloss”: Multiclass logloss
  • “auc”: Area under the curve for ranking evaluation.
  • “ndcg”:Normalized Discounted Cumulative Gain
  • “map”:Mean average precision
  • “ndcg@n”,”map@n”: n can be assigned as an integer to cut off the top positions in the lists for evaluation.
  • “ndcg-“,”map-“,”ndcg@n-“,”map@n-“: In XGBoost, NDCG and MAP will evaluate the score of a list without any positive samples as 1. By adding “-” in the evaluation metric XGBoost will evaluate these score as 0 to be consistent under some conditions.

(3)lambda [default=0] L2 正则的惩罚系数

(4)alpha [default=0] L1 正则的惩罚系数

(5)lambda_bias 在偏置上的L2正则。缺省值为0(在L1上没有偏置项的正则,因为L1时偏置不重要)

(6)eta [default=0.3] 
为了防止过拟合,更新过程中用到的收缩步长。在每次提升计算之后,算法会直接获得新特征的权重。 eta通过缩减特征的权重使提升计算过程更加保守。缺省值为0.3 
取值范围为:[0,1]

(7)max_depth [default=6] 数的最大深度。缺省值为6 ,取值范围为:[1,∞]

(8)min_child_weight [default=1] 
孩子节点中最小的样本权重和。如果一个叶子节点的样本权重和小于min_child_weight则拆分过程结束。在现行回归模型中,这个参数是指建立每个模型所需要的最小样本数。该成熟越大算法越conservative 
取值范围为: [0,∞]

xgb1=XGBClassifier(
learning_rate=0.1,
n_estimators=1000,
max_depth=5,
min_child_weight=1,
gamma=0,
subsample=0.8
colsample_bytree=0.8,
objective='binary:logistic',
nthread=4,
scale_pos_weight=1,
seed=27)

3、xgboost重要模块:plot_importance【显示特征的重要性】

from xgboost import XGBClassifier
from xgboost import plot_importance
from matplotlib import pyplot model=XGBClassifier()
model.fit(X,Y)
plot_importance(model)
pyplot.show()
#图中就可以显示出各种特征的重要性

xgboost学习的更多相关文章

  1. 【新人赛】阿里云恶意程序检测 -- 实践记录11.10 - XGBoost学习 / 代码阅读、调参经验总结

    XGBoost学习: 集成学习将多个弱学习器结合起来,优势互补,可以达到强学习器的效果.要想得到最好的集成效果,这些弱学习器应当"好而不同". 根据个体学习器的生成方法,集成学习方 ...

  2. XGboost学习总结

    XGboost,全称Extrem Gradient boost,极度梯度提升,是陈天奇大牛在GBDT等传统Boosting算法的基础上重新优化形成的,是Kaggle竞赛的必杀神器. XGboost属于 ...

  3. xgboost学习与总结

    最近在研究xgboost,把一些xgboost的知识总结一下.这里只是把相关资源作总结,原创的东西不多. 原理 xgboost的原理首先看xgboost的作者陈天奇的ppt 英文不太好的同学可以看看这 ...

  4. 数据竞赛利器 —— xgboost 学习清单

    1. 入门大全 xgboost 作者给出的一份完备的使用 xgboost 进行数据分析的完整示例代码:A walk through python example for UCI Mushroom da ...

  5. XGBoost学习笔记2

    XGBoost API 参数 分类问题 使用逻辑回归 # Import xgboost import xgboost as xgb # Create arrays for the features a ...

  6. XGBoost学习笔记1

    XGBoost XGBoost这个网红大杀器,似乎很好用,完事儿还是自己推导一遍吧,datacamp上面有辅助的课程,但是不太涉及原理 它究竟有多好用呢?我还没用过,先搞清楚原理,hahaha~ 参考 ...

  7. 【Python机器学习实战】决策树与集成学习(七)——集成学习(5)XGBoost实例及调参

    上一节对XGBoost算法的原理和过程进行了描述,XGBoost在算法优化方面主要在原损失函数中加入了正则项,同时将损失函数的二阶泰勒展开近似展开代替残差(事实上在GBDT中叶子结点的最优值求解也是使 ...

  8. xgboost原理及应用

    1.背景 关于xgboost的原理网络上的资源很少,大多数还停留在应用层面,本文通过学习陈天奇博士的PPT 地址和xgboost导读和实战 地址,希望对xgboost原理进行深入理解. 2.xgboo ...

  9. xgboost原理及应用--转

    1.背景 关于xgboost的原理网络上的资源很少,大多数还停留在应用层面,本文通过学习陈天奇博士的PPT地址和xgboost导读和实战 地址,希望对xgboost原理进行深入理解. 2.xgboos ...

随机推荐

  1. WEBGL学习【四】模型视图矩阵

    <html lang="zh-CN"> <!--服务器运行地址:http://127.0.0.1:8080/webgl/LearnNeHeWebGL/NeHeWe ...

  2. 【SPOJ 104】HIGH - Highways (高斯消元)

    题目描述 In some countries building highways takes a lot of time- Maybe that's because there are many po ...

  3. docker数据卷的使用 -v --volumes--from

    总结一下docker数据管理的三种方法: 1.普通的挂在数据: -v docker run  -v /father/path:/child/path-v 参数会把当前系统的文件目录/father/pa ...

  4. 一次 Laravel 性能分析全程笔记

    大家都知道 laravel 项目写起来是挺爽,但是在生产环境性能不高,我们来抽丝剥茧分析我自己项目的运行时间消耗: Bootstrap 耗时 步骤 耗时 Illuminate\Foundation\B ...

  5. android生成sdk.jar 小工具

    net.sf.fjep.fatjar_0.0.31.jar 生成jar的工具 把这个net.sf.fjep.fatjar_0.0.31.jar到Eclipse的plugins中,从启Eclipse.点 ...

  6. Zepto.js实现fadeIn,fadeOut功能

    Zepto是一个轻量级的针对现代高级浏览器的JavaScript库, 它与jquery有着类似的api. 如果你会用jquery,那么你也会用zepto. Zepto的设计目的是提供 jQuery 的 ...

  7. jquery-ajax基础-XMLHttpRequest

    XMLHttpRequest知识点 原生的ajax代码 var xmlhttp; // 声明一个对象 if (window.XMLHttpRequest) {// code for IE7+, Fir ...

  8. HDU 2815

    特判B不能大于等于C 高次同余方程 #include <iostream> #include <cstdio> #include <cstring> #includ ...

  9. JavaWeb应用中的身份验证(声明式)——基于表单的身份认证

    容器管理安全最普遍的类型建立在基于表单的身份验证方式上. 通过这样的方式,server自己主动将尚未验证的用户重定向到一个HTML表单.检查他们的username和password,决定他们属于哪个角 ...

  10. FZU Problem 1853 Number Deletion

    Problem 1853 Number Deletion Accept: 80    Submit: 239 Time Limit: 1000 mSec    Memory Limit : 32768 ...