使用tensorflow编写MLP分类器 - 机器学习和生物信息学实验室联盟

import numpy as np
import tensorflow as tf
from sklearn.metrics import roc_auc_score
from sklearn import metrics
from sklearn.cross_validation import KFold
def dense_to_one_hot(labels_dense,num_classes=2):
""" convert class lables from scalars to one-hot vector"""
labels_dense = np.asarray(labels_dense)
num_labels = labels_dense.shape[0]
index_offset = np.arange(num_labels)*num_classes
labels_one_hot = np.zeros((num_labels, num_classes))
labels_one_hot.flat[index_offset + labels_dense.ravel()] = 1
return labels_one_hot
def kfold(trainData,trainClass,nFold=10):
skf = KFold(len(trainData),nFold,shuffle=True,random_state=1234)
kDataTrain = []
kDataTrainC = []
kDataTest = []
kDataTestC = []
trainData = np.asarray(trainData)
trainClass = np.asarray(trainClass)
for train_index,test_index in skf:
X_train,X_test = trainData[train_index],trainData[test_index]
y_train,y_test = trainClass[train_index],trainClass[test_index]
kDataTrain.append(X_train)
kDataTrainC.append(y_train)
kDataTest.append(X_test)
kDataTestC.append(y_test)
return kDataTrain,kDataTrainC,kDataTest,kDataTestC
def load_data(fileName):
lables = []
feature = []
for line in open(fileName):
if line.startswith("@") or line == "":
continue
listV = line.strip().split(",")
feature.append(listV[0:-2])
lables.append(int(listV[-1]))
return lables,feature
#return dense_to_one_hot(lables),np.asarray(feature)
ty,tx = load_data("feature_91.arff")
kDataTrain,kDataTrainC,kDataTest,kDataTestC = kfold(tx,ty)
acc =[]
for index in range(len(kDataTrain)):
print "cross validation:",index
ty,tx = kDataTrainC[index],kDataTrain[index]
testy,testx = kDataTestC[index],kDataTest[index]
ty = dense_to_one_hot(ty)
testy = dense_to_one_hot(testy)
learning_rate = 0.0005
training_epochs = 500
batch_size = 100
n_hidden_1 = 300
n_hidden_2 = 300
n_input = tx.shape[1]
n_class = 2
x = tf.placeholder("float",[None,n_input])
y = tf.placeholder("float",[None,n_class])
def mlp(x,weights,biases):
layer_1 = tf.add(tf.matmul(x,weights["h1"]),biases["b1"])
layer_1 = tf.nn.relu(layer_1)
layer_2 = tf.add(tf.matmul(layer_1, weights["h2"]), biases["b2"])
layer_2 = tf.nn.relu(layer_2)
out_layer = tf.matmul(layer_2,weights['out']) + biases['out']
return out_layer
weights = {
'h1': tf.Variable(tf.random_normal([n_input,n_hidden_1])),
'h2': tf.Variable(tf.random_normal([n_hidden_1,n_hidden_2])),
'out': tf.Variable(tf.random_normal([n_hidden_2,n_class]))
}
biases = {
'b1': tf.Variable(tf.random_normal([n_hidden_1])),
'b2': tf.Variable(tf.random_normal([n_hidden_2])),
'out':tf.Variable(tf.random_normal([n_class]))
}
pred = mlp(x,weights,biases)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(pred,y))
optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)
init = tf.initialize_all_variables()
with tf.Session() as sess:
sess.run(init)
for i in range(training_epochs):
avg_cost = 0.
total_batch = int(tx.shape[0]/batch_size)
for start,end in zip(range(0,len(tx),batch_size), range(batch_size,len(tx),batch_size)):
_,loss = sess.run([optimizer,cost],feed_dict={x:tx[start:end],y:ty[start:end]})
avg_cost += loss / total_batch
#print i,"loss:",avg_cost
correct_prediction = tf.equal(tf.argmax(pred, 1), tf.argmax(testy, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
result = accuracy.eval({x: testx, y: testy})
acc.append(result)
print "Accuracy:", result
print "cross validation result"
print "accuracy:",np.mean(acc)

复制代码