neural network simple model

simple_neuralnet.py

import tensorflow as tf
import matplotlib.pyplot as plt
%matplotlib inline

samples = 1000
up = [i for i in range(10)]
down = [9-i for i in range(10)]
data = [up if i%2 == 0 else down for i in range(samples)]
label = [[1] if i%2 == 0 else [0] for i in range(samples)]

print("data: ", data[:10])
print()
print("label: ", label[:10])

x = tf.placeholder(tf.float32, shape=[None, 10])
y_ = tf.placeholder(tf.float32)

layer1 = tf.layers.dense(x, 20)
layer2 = tf.layers.dense(layer1, 15)
layer3 = tf.layers.dense(layer2, 30)
layer4 = tf.layers.dense(layer3, 10)
layer5 = tf.layers.dense(layer4, 5)
out = tf.layers.dense(layer5, 1)

loss = tf.losses.sigmoid_cross_entropy(y_, out)
train_op = tf.train.GradientDescentOptimizer(1e-2).minimize(loss)

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    for i in range(100):
        _, _loss = sess.run([train_op, loss], {x: data, y_: label})
        print("step: {}, loss: {}".format(i, _loss))
    
    pred = sess.run(tf.nn.sigmoid(out), {x: data[:10]})
    print()
    print("prediction: {}".format(pred))