''' Created on 2017年5月21日 @author: weizhen ''' #Tensorflow的另外一个高层封装TFLearn(集成在tf.contrib.learn里)对训练Tensorflow模型进行了一些封装 #使其更便于使用。 #使用TFLearn实现分类问题 #为了方便数据处理，本程序使用了sklearn工具包， #更多信息可以参考http://scikit-learn.org from sklearn import model_selection from sklearn import datasets from sklearn import metrics import tensorflow as tf import numpy as np from tensorflow.contrib.learn.python.learn.estimators.estimator import SKCompat #导入TFLearn learn = tf.contrib.learn #自定义模型，对于给定的输入数据(features)以及其对应的正确答案(target) #返回在这些输入上的预测值、损失值以及训练步骤 def my_model(features,target): #将预测的目标转换为one-hot编码的形式，因为共有三个类别，所以向量长度为3.经过转化后，第一个类别表示为(1,0,0) #第二个为(0,1,0)，第三个为(0,0,1) target = tf.one_hot(target,3,1,0) #定义模型以及其在给定数据上的损失函数 logits = tf.contrib.layers.fully_connected(features,3,tf.nn.softmax) loss = tf.losses.softmax_cross_entropy(target, logits) #创建模型的优化器，并得到优化步骤 train_op=tf.contrib.layers.optimize_loss(loss, #损失函数 tf.contrib.framework.get_global_step(), #获取训练步数并在训练时更新 optimizer='Adam', #定义优化器 learning_rate=0.01) #定义学习率 #返回在给定数据上的预测结果、损失值以及优化步骤 return tf.arg_max(logits, 1),loss,train_op #加载iris数据集，并划分为训练集合和测试集合 iris = datasets.load_iris() x_train,x_test,y_train,y_test=model_selection.train_test_split(iris.data, iris.target, test_size=0.2, random_state=0) #将数据转化为float32格式 x_train,x_test = map(np.float32,[x_train,x_test]) #封装和训练模型，输出准确率 classifier=SKCompat(learn.Estimator(model_fn=my_model,model_dir="Models/model_1")) #使用封装好的模型和训练数据执行100轮迭代 classifier.fit(x_train,y_train,steps=800) #使用训练好的模型进行结果预测 y_predicted=[i for i in classifier.predict(x_test)] #计算模型的准确度 score=metrics.accuracy_score(y_test,y_predicted) print("Accuracy: %.2f"%(score*100))
结果如下所示
2017-05-21 15:49:11.386435: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.386846: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE2 instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.387271: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.387604: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.388450: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.388882: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.389180: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations. 2017-05-21 15:49:11.389766: W c: f_jenkinshomeworkspace elease-windevicecpuoswindows ensorflowcoreplatformcpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations. Accuracy: 100.00