tensorflow-loss

#define input number is 1,output num is 1
x=tf.placeholder(float,"input")
y=tf.placeholder(float,"output")

w=tf.Variable(0.0)
b=tf.Variable(0.0)
y1=tf.matmul(x,w)+b
loss=tf.square(y1-y,"loss")

#define input number is n,features num is f,output num is 1
x=tf.placeholder(float,name="input",shape=[n,f])
y=tf.placeholder(float,name="output")

w=tf.Variable(tf.random_normal([f,1]))
b=tf.Variable(0.0)
y1=tf.matmul(x,w)+b
loss=tf.reduce_mean(tf.square(y1-y,"loss"))

#define input number is n,features num is f,output num is o
x=tf.placeholder(float,name="input",shape=[n,f])
y=tf.placeholder(float,1,name="output",shape=[1,o])

w=tf.Variable(tf.random_normal([f,1]),name="weights")
b=tf.Variable(tf.random_normal([1,o]),name="bias")
y1=tf.matmul(x,w)+b

entropy=tf.nn.softmax_cross_entropy_with_logits(y1,y)
loss=tf.reduce_mean(entropy)
----------------------------------------------------------------------------
#根据需要选择下列之一

// L1 正则化
lamda=tf.constant(0.8)
L1=lamda*tf.reduce_sum(tf.abs(w))
loss+=L1

//L2正则化
lamda=tf.constant(0.8)
L2=lamda*tf.nn.l2_loss(w)
loss+=L2