Learning with ensembles
来源:互联网 发布:rng淘宝官方旗舰店 编辑:程序博客网 时间:2024/05/22 15:55
The goal behind ensemble methodsis to combine different classifers into a meta-classifer that has a better generalization performance than each individual classifer alone。focus on the most popular ensemble methods that use the majority votingprinciple。
1. Ensemble error
from scipy.misc import combimport mathdef ensemble_error(n_classifier, error): k_start = math.ceil(n_classifier / 2.0) probs = [comb(n_classifier, k) * error**k * (1 - error)**(n_classifier - k) for k in range(int(k_start), n_classifier + 1)] return sum(probs)print('Ensemble error', ensemble_error(n_classifier=11, error=0.25))('Ensemble error', 0.034327507019042969)
2. compute the ensemble error rates for a range of different base errors from 0.0 to 1.0 to visualize the relationship between ensemble and base errors in a line graph
import numpy as nperror_range = np.arange(0.0, 1.01, 0.01)ens_errors = [ensemble_error(n_classifier=11, error=error) for error in error_range]plt.plot(error_range, ens_errors, label='Ensemble error', linewidth=2)plt.plot(error_range, error_range, linestyle='--', label='Base error', linewidth=2)plt.xlabel('Base error')plt.ylabel('Base/Ensemble error')plt.legend(loc='upper left')plt.grid()plt.show()
As we can see in the resulting plot, the error probability of an ensemble is always better than the error of an individual base classifer as long as the base classifers perform better than random guessing (ε <0.5 )
Reference:《Python Machine Learning》
0 0
- Learning with ensembles
- Feature transformations with ensembles of trees
- learning with errors理念
- Learning C with gdb
- Machine Learning with OpenCV
- Deep Learning with Torch
- deep learning with R
- Multiclass Learning with ECOC
- Learning Javascript with Object Graphs
- Learning D3.js with App
- Self Learning with Sed<1>
- Learning Time Series with R
- Deep Learning algorithms with TensorFlow
- machine learning with spark (1)
- Machine Learning With Spark--读书笔记
- learning R with swirl-vectors
- learning R with swirl-logic
- learning R with swirl-functions
- HDU 3068 最长回文
- 题目1201:二叉排序树
- 自己整理的eclipse的快捷键
- strsep的作用
- web worker
- Learning with ensembles
- VS2013创建CUDA文件
- LeetCode 217. Contains Duplicate
- 【java】二叉树重建
- 高精度开方
- 列表渲染多层嵌套循环及wx:key的使用
- excle+pytest+jenkins+allure接口自动化测试框架
- 题目1202:排序
- oracle11g active dataguard dupliacte制作从库