240716 svm找到最优参数

结果

代码

# -*- coding: utf-8 -*-
import numpy as np
import matplotlib.pyplot as plt
from sklearn import svm, grid_search, cross_validation
from sklearn.metrics import classification_report

import utilities 

# 加载数据
input_file = 'data_multivar.txt'
X, y = utilities.load_data(input_file)

###############################################
# 切分出测试数据

X_train, X_test, y_train, y_test = cross_validation.train_test_split(X, y, test_size=0.25, random_state=5)

# 设置方法
parameter_grid = [  {'kernel': ['linear'], 'C': [1, 10, 50, 600]},
                    {'kernel': ['poly'], 'degree': [2, 3]},
                    {'kernel': ['rbf'], 'gamma': [0.01, 0.001], 'C': [1, 10, 50, 600]},
                 ]

metrics = ['precision', 'recall_weighted']
#打印结果
for metric in metrics:
    print "\n#### Searching optimal hyperparameters for", metric

    classifier = grid_search.GridSearchCV(svm.SVC(C=1), 
            parameter_grid, cv=5, scoring=metric)
    classifier.fit(X_train, y_train)
# 评价分数
    print "\nScores across the parameter grid:"
    for params, avg_score, _ in classifier.grid_scores_:
        print params, '-->', round(avg_score, 3)

    print "\nHighest scoring parameter set:", classifier.best_params_

    y_true, y_pred = y_test, classifier.predict(X_test)
    print "\nFull performance report:\n"
    print classification_report(y_true, y_pred)