LibSVM 是SVM分类和回归问题的有效解决方案。这里的以php接口封装的svm扩展是为了在php脚本中更方便的应用 支持向量机(*SVM*)号称最优秀的分类算法之一 [https://pecl.php.net/package/svm](https://pecl.php.net/package/svm) * [例子](https://www.php.net/manual/en/svm.examples.php) * [SVM](https://www.php.net/manual/en/class.svm.php)— SVM类 * [SVM :: \_\_ construct](https://www.php.net/manual/en/svm.construct.php)—构造一个新的SVM对象 * [SVM :: crossvalidate](https://www.php.net/manual/en/svm.crossvalidate.php)—在训练数据的子集上测试训练参数 * [SVM :: getOptions](https://www.php.net/manual/en/svm.getoptions.php)—返回当前的训练参数 * [SVM :: setOptions](https://www.php.net/manual/en/svm.setoptions.php)—设置训练参数 * [SVM :: train](https://www.php.net/manual/en/svm.train.php)—基于训练数据创建SVMModel * [SVMModel](https://www.php.net/manual/en/class.svmmodel.php)— SVMModel类 * [SVMModel :: checkProbabilityModel](https://www.php.net/manual/en/svmmodel.checkprobabilitymodel.php)—如果模型具有概率信息，则返回true * [SVMModel :: \_\_ construct](https://www.php.net/manual/en/svmmodel.construct.php)—构造一个新的SVMModel * [SVMModel :: getLabels](https://www.php.net/manual/en/svmmodel.getlabels.php)—获取在其上训练模型的标签 * [SVMModel :: getNrClass](https://www.php.net/manual/en/svmmodel.getnrclass.php)—返回使用模型训练的类的数量 * [SVMModel :: getSvmType](https://www.php.net/manual/en/svmmodel.getsvmtype.php)—获取模型训练所使用的SVM类型 * [SVMModel :: getSvrProbability](https://www.php.net/manual/en/svmmodel.getsvrprobability.php)—获取回归类型的sigma值 * [SVMModel :: load](https://www.php.net/manual/en/svmmodel.load.php)—加载保存的SVM模型 * [SVMModel :: predict\_probability](https://www.php.net/manual/en/svmmodel.predict-probability.php)—先前看不见的数据的返回类概率 * [SVMModel :: predict](https://www.php.net/manual/en/svmmodel.predict.php)—预测以前看不见的数据的值 * [SVMModel :: save](https://www.php.net/manual/en/svmmodel.save.php)—将模型保存到文件