sudo pip install numpy

sudo apt-get install python-matplotlib ipython ipython-notebook
sudo apt-get install python-pandas python-sympy python-nose
sudo pip install scipy

sudo pip install -U scikit-learn

pip list

import numpy
import scipy
import sklearn

import numpy as np
feature_matrix = np.load('dataSet/feature_matrix_save.npy')
class_list = np.load('dataSet/class_result_save.npy')

from sklearn import preprocessing
# 归一化（Normalization）
normalized_X = preprocessing.normalize(feature_matrix)
print normalized_X
# 标准化（Standardization）
standardized_X = preprocessing.scale(feature_matrix)
print standardized_X

from sklearn.ensemble import ExtraTreesClassifier
model = ExtraTreesClassifier()
print feature_matrix.shape # 原特征矩阵规模
feature_matrix = model.fit(feature_matrix, class_list).transform(feature_matrix)
print feature_matrix.shape # 特征选择后 特征矩阵的规模

from sklearn.feature_extraction.text import TfidfTransformer
tfidf_transformer = TfidfTransformer()
feature_matrix = tfidf_transformer.fit_transform(feature_matrix).toarray()

from sklearn import metrics
from sklearn.naive_bayes import GaussianNB
# 构建朴素贝叶斯模型
model = GaussianNB()
model.fit(feature_matrix, class_list)
print model
# 使用测试集进行测试(此处将训练集做测试集)
expected = class_list
predicted = model.predict(feature_matrix)
# 输出测试效果
print metrics.classification_report(expected, predicted)
print metrics.confusion_matrix(expected, predicted)

from sklearn import metrics
from sklearn.neighbors import KNeighborsClassifier
# 构建knn模型
model = KNeighborsClassifier()
model.fit(feature_matrix, class_list)
print model
# 使用测试集进行测试(此处将训练集做测试集)
expected = class_list
predicted = model.predict(feature_matrix)
# 输出测试效果
print metrics.classification_report(expected, predicted)
print metrics.confusion_matrix(expected, predicted)

from sklearn import metrics
from sklearn.tree import DecisionTreeClassifier
# 构建决策数模型
model = DecisionTreeClassifier()
model.fit(feature_matrix, class_list)
print model
# 使用测试集进行测试(此处将训练集做测试集)
expected = class_list
predicted = model.predict(feature_matrix)
# 输出测试效果
print metrics.classification_report(expected, predicted)
print metrics.confusion_matrix(expected, predicted)