Python机器学习库scikit-learn

概述

scikit-learn 是机器学习领域非常热门的一个开源库，基于Python 语言写成。可以免费使用。 而且使用非常的简单，文档感人，非常值得去学习。

下面是一张scikit-learn的图谱：

我们可以看到，机器学习分为四大块，分别是 classification (分类)， clustering (聚类), regression (回归), dimensionality reduction (降维)。

安装scikit-learn

如果使用的是ubuntu则非常的简单，直接sudo apt-get install scikit-learn即可，这里可能会有要你安装别的依赖，也是同样的安装方法，如果是别的linux版本，可使用pip等工具进行安装。

测试：

# 不报错则表示安装成功
>>> import sklearn
>>> 

安装XGBDT

本质上还是GBDT，只是对GBDT进行了一些更改，叫X (Extreme) GBoosted，它把速度和效率做到了极致。在scikit-learn目前还没有这个分类器，因此要进行单独的安装。

这里对linux的安装进行说明，其余操作系统见：
http://xgboost.readthedocs.io/en/latest/build.html

# 拉取源码包
git clone --recursive https://github.com/dmlc/xgboost
cd xgboost
# 编译
make -j4

# python包的安装
# 首先安装工具
sudo apt-get install python-setuptools

#　进入目录，安装
cd python-package
sudo python setup.py install
#　测试不报错，成功
>>> import xgboost
>>> 

scikit-learn测试

测试的数据为，美国一个区域的糖尿病的情况，具有以下的信息：

Attribute Information:
1. Number of times pregnant
2. Plasma glucose concentration a 2 hours in an oral glucose tolerance test
3. Diastolic blood pressure (mm Hg)
4. Triceps skin fold thickness (mm)
5. 2-Hour serum insulin (mu U/ml)
6. Body mass index (weight in kg/(height in m)^2)
7. Diabetes pedigree function
8. Age (years)
9. Class variable (0 or 1)

第9个是标签，即我们要预测的情况，0表示没有患病，1表示患病，数据集下载地址：
https://archive.ics.uci.edu/ml/datasets/Pima+Indians+Diabetes

完整代码如下，可以进行一键测试多个算法：

#!usr/bin/env python  
#-*- coding: utf-8 -*-  

import time  
from sklearn import metrics  
import numpy as np  
from numpy import *
from sklearn import cross_validation

# Multinomial Naive Bayes Classifier  
def naive_bayes_classifier(train_x, train_y):  
    from sklearn.naive_bayes import MultinomialNB  
    model = MultinomialNB(alpha=0.01)  
    model.fit(train_x, train_y)  
    return model  


# KNN Classifier  
def knn_classifier(train_x, train_y):  
    from sklearn.neighbors import KNeighborsClassifier  
    model = KNeighborsClassifier(n_neighbors=10)  
    model.fit(train_x, train_y)  
    return model  


# Logistic Regression Classifier  
def logistic_regression_classifier(train_x, train_y):  
    from sklearn.linear_model import LogisticRegression  
    model = LogisticRegression(penalty='l2')  
    model.fit(train_x, train_y)  
    return model  


# Random Forest Classifier  
def random_forest_classifier(train_x, train_y):  
    from sklearn.ensemble import RandomForestClassifier  
    model = RandomForestClassifier(n_estimators=100)  
    model.fit(train_x, train_y)  
    return model  


# Decision Tree Classifier  
def decision_tree_classifier(train_x, train_y):  
    from sklearn import tree  
    model = tree.DecisionTreeClassifier()  
    model.fit(train_x, train_y)  
    return model  


# GBDT(Gradient Boosting Decision Tree) Classifier  
def gradient_boosting_classifier(train_x, train_y):  
    from sklearn.ensemble import GradientBoostingClassifier  
    model = GradientBoostingClassifier(n_estimators=40)  
    model.fit(train_x, train_y)  
    return model  


# SVM Classifier  
def svm_classifier(train_x, train_y):  
    from sklearn.svm import SVC  
    model = SVC(kernel='rbf', probability=True)  
    model.fit(train_x, train_y)  
    return model  

# SVM Classifier using cross validation  
def svm_cross_validation(train_x, train_y):  
    from sklearn.grid_search import GridSearchCV  
    from sklearn.svm import SVC  
    model = SVC(kernel='rbf', probability=True)  
    param_grid = {'C': [1e-3, 1e-2, 1e-1, 1, 10, 100, 1000], 'gamma': [0.001, 0.0001]}  
    grid_search = GridSearchCV(model, param_grid, n_jobs = 1, verbose=1)  
    grid_search.fit(train_x, train_y)  
    best_parameters = grid_search.best_estimator_.get_params()  
    for para, val in best_parameters.items():  
        print para, val  
    model = SVC(kernel='rbf', C=best_parameters['C'], gamma=best_parameters['gamma'], probability=True)  
    model.fit(train_x, train_y)  
    return model  

# XGBoost Classfier
def extreme_gradient_boosting_classifier(train_x,train_y):
    import xgboost
    model = xgboost.XGBClassifier()
    model.fit(train_x,train_y)
    return model

# read dataset
def read_data():  
    dataset = np.loadtxt('diabetes.txt',delimiter=',')
    x = dataset[:,:8]
    y = dataset[:,8]
    seed = 7
    test_size = 0.33
    # split the dataset 
    train_x,test_x,train_y,test_y = cross_validation.train_test_split \
        (x,y,test_size=test_size,random_state=seed)
    return train_x, test_x, train_y, test_y  


if __name__ == '__main__':        
    test_classifiers = ['NB','RF','SVM','KNN','LR','DT','GBDT','XGBDT']
    classifiers = {'NB':naive_bayes_classifier,   
                  'KNN':knn_classifier,  
                   'LR':logistic_regression_classifier,  
                   'RF':random_forest_classifier,  
                   'DT':decision_tree_classifier,  
                  'SVM':svm_classifier,  
                'SVMCV':svm_cross_validation,  
                 'GBDT':gradient_boosting_classifier,  
                'XGBDT':extreme_gradient_boosting_classifier
    }  

    print 'reading training and testing data...'  
    train_x, test_x, train_y, test_y = read_data()  
    num_train, num_feat = train_x.shape  
    num_test, num_feat = test_x.shape  
    is_binary_class = (len(np.unique(train_y)) == 2)  
    print '******************** Data Info *********************'  
    print '#training data: %d, #testing_data: %d, dimension: %d' % (num_train, num_test, num_feat)  

    for classifier in test_classifiers:  
        print '******************* %s ********************' % classifier  
        start_time = time.time()  
        model = classifiers[classifier](train_x, train_y)  
        print 'training took %fs!' % (time.time() - start_time)  
        predict = model.predict(test_x)  
        if is_binary_class:  
            precision = metrics.precision_score(test_y, predict)  
            recall = metrics.recall_score(test_y, predict)  
            print 'precision: %.2f%%, recall: %.2f%%' % (100 * precision, 100 * recall)  
        accuracy = metrics.accuracy_score(test_y, predict)  
        print 'accuracy: %.2f%%' % (100 * accuracy)   

运行结果，可以看到在此问题上XGBDT效果稍微好一点，但是这是没有经过调参的，可以进行调参等预处理操作来改善效果。

yqtao@yqtao:~/machine-learn$ python sklean.py 
reading training and testing data...
******************** Data Info *********************
#training data: 514, #testing_data: 254, dimension: 8
******************* NB ********************
training took 0.001360s!
precision: 48.35%, recall: 47.83%
accuracy: 62.60%
******************* RF ********************
training took 0.198425s!
precision: 71.26%, recall: 67.39%
accuracy: 78.35%
******************* SVMCV ********************
Fitting 3 folds for each of 14 candidates, totalling 42 fits
[Parallel(n_jobs=1)]: Done  42 out of  42 | elapsed:    1.5s finished
kernel rbf
C 1
verbose False
probability True
degree 3
shrinking True
max_iter -1
decision_function_shape None
random_state None
tol 0.001
cache_size 200
coef0 0.0
gamma 0.0001
class_weight None
training took 1.536800s!
precision: 69.84%, recall: 47.83%
accuracy: 73.62%
******************* KNN ********************
training took 0.003870s!
precision: 71.21%, recall: 51.09%
accuracy: 74.80%
******************* LR ********************
training took 0.003629s!
precision: 70.83%, recall: 55.43%
accuracy: 75.59%
******************* DT ********************
training took 0.002498s!
precision: 61.18%, recall: 56.52%
accuracy: 71.26%
******************* GBDT ********************
training took 0.033451s!
precision: 70.73%, recall: 63.04%
accuracy: 77.17%
******************* XGBDT ********************
training took 0.232969s!
precision: 70.45%, recall: 67.39%
accuracy: 77.95%

参考资料：

http://scikit-learn.org/stable/
http://blog.csdn.net/zouxy09/article/details/48903179
http://blog.csdn.net/matrix_space/article/details/50541217