x = self.pool(F.relu(self.conv2(x)))
x = self.pool(F.relu(self.conv3(x)))
x = x.reshape(-1, self.get_features(x)) # 此步必须有
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = F.relu(self.fc3(x))
return x
if __name__ == '__main__':
net = Net()
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.001)
epochs = 100
batch_size = 4000
train_data, train_label = getData.getTrainData()
t = 0
while t < 6:
acc = 0.0
los = 0.0
inputs, target = train_data, train_label
for i in range(batch_size):
net.zero_grad()
outputs = net(inputs[i])
loss = criterion(outputs, target[i].long())
loss.backward()
optimizer.step()
los += loss
predict = 0
for j in range(15):
if outputs[0][j] == max(outputs[0]):
outFile.write(','+str(i))
outFile.write("\n")
if __name__ == '__main__':
start = time.clock()
nowtime,lastmonth=getTimes()
nowtime='2014-5-1'
lastmonth='2014-4-1'
R_Num=8 #推荐职位数
#获取数据
allJob_SN=getAllJob_SN(nowtime,lastmonth)#时间段内所有职位
Jw_SN=getAllJw_SN()#获得所有求职者
JWINFO,JOB_OFFER=getJWJOBINFO(nowtime,lastmonth)#获得求职者信息及职位信息
trainData=getTrainData(nowtime,lastmonth)
getData_time=time.clock()
print u"getData耗时: %f s" % (getData_time - start)
#userCF_IIF推荐
finallyRecommend=userCF_IIF_finallyRecommend(trainData,JWINFO,JOB_OFFER,R_Num)
userCF_IIF_time=time.clock()
print u"userCF_IIF耗时: %f s" % (userCF_IIF_time - getData_time)
#itemCF_IUF推荐
finallyRecommend=itemCF_IUF_finallyRecommend(trainData,JWINFO,JOB_OFFER,R_Num,finallyRecommend)
itemCF_IUF_time=time.clock()
print u"itemCF_IUF耗时: %f s" % (itemCF_IUF_time - userCF_IIF_time)
#city_type_most_popular推荐
outFile.write("\n")
if __name__ == '__main__':
start = time.clock()
nowtime, lastmonth = getTimes()
nowtime = '2016-7-28'
lastmonth = '2016-6-1'
R_Num = 8 #推荐职位数
#获取数据
Job_SN = getAllJob_SN(nowtime, lastmonth) #时间段内所有职位
Jw_SN = getAllJw_SN() #获得所有求职者
JWINFO, JOB_OFFER = getJWJOBINFO(nowtime, lastmonth) #获得求职者信息及职位信息
trainData = getTrainData(nowtime, lastmonth)
getData_time = time.clock()
print u"getData耗时: %f s" % (getData_time - start)
#userCF_IIF推荐
finallyRecommend = userCF_IIF_finallyRecommend(trainData, JWINFO,
JOB_OFFER, R_Num)
userCF_IIF_time = time.clock()
print u"userCF_IIF耗时: %f s" % (userCF_IIF_time - getData_time)
#itemCF_IUF推荐
finallyRecommend = itemCF_IUF_finallyRecommend(trainData, JWINFO,
JOB_OFFER, R_Num,
finallyRecommend)
itemCF_IUF_time = time.clock()