这几天在B站看莫烦的视频,学习一波,给出视频地址:https://www.bilibili.com/video/av16001891/?p=22
先放出代码
#####搭建神经网络测试 def add_layer(inputs,in_size,out_size,activation_function=None): Weights = tf.Variable(tf.random_normal([in_size, out_size],dtype=np.float32)) biases = tf.Variable(tf.zeros([1,out_size])+0.1) Wx_plus_b = tf.matmul(inputs, Weights)+biases if activation_function is None: outputs = Wx_plus_b else: outputs = activation_function(Wx_plus_b) return outputs x_data = np.linspace(-1,1,300)[:, np.newaxis] noise = np.random.normal(0,0.05,x_data.shape) y_data = np.square(x_data)-0.5+noise xs = tf.placeholder(tf.float32,[None,1]) ys = tf.placeholder(tf.float32,[None,1]) l1 = add_layer(xs,1,10,activation_function=tf.nn.relu) prediction = add_layer(l1,10,1,activation_function=None) loss = tf.reduce_mean(tf.reduce_sum(tf.square(ys - prediction), reduction_indices=[1])) train_step = tf.train.GradientDescentOptimizer(0.1).minimize(loss) init = tf.global_variables_initializer() with tf.Session() as sess: sess.run(init) for i in range(1000): sess.run(train_step,feed_dict={xs:x_data,ys:y_data}) if i% 50 ==0: print(sess.run(loss,feed_dict={xs:x_data,ys:y_data})) #####
首先,在add_layer函数中,参数有inputs,in_size,out_size,activation_function=None
其中inupts是输入,in_size是输入维度,out_size是输出维度, activation_function是激活函数,
Weights是权重,维度是(in_size*out_size);
bias是偏置,维度是(1*out_size);
Wx_plus_b的维度和out_size相同;
x_data = np.linspace(-1,1,300)[:, np.newaxis]这步操作,表示生成-1到1之间均匀分布的300个数,然后转换维度,变成(300,1);noise和y_data的维度均和
x_data相同;
xs = tf.placeholder(tf.float32,[None,1])和ys = tf.placeholder(tf.float32,[None,1])表示生成xs和ys变量的占位符,维度是(None,1),不知道有多少行,但只要1列;
l1 = add_layer(xs,1,10,activation_function=tf.nn.relu)表示xs是inputs,in_size是1,out_size是10,激活函数是relu;添加了一层神经网络
prediction = add_layer(l1,10,1,activation_function=None)表示输入是l1,in_size是10,out_size是1,没有激活函数
接下去是计算损失,loss = tf.reduce_mean(tf.reduce_sum(tf.square(ys - prediction),reduction_indices=[1]))
之后一步是用梯度下降来优化损失函数;
解释一下为什么不直接在add_layer函数中使用x_data:x_data是ndarray格式,Weights是Variable格式,不能直接相乘,所以要在session会话中用字典格式传入x_data和y_data, 也就是sess.run(train_step,feed_dict={xs:x_data,ys:y_data})