3.1 tensorflow简单线性回归

Linear Regression线性回归

线性回归也即最小二乘,就是用线性函数 y=Ax+b 去拟合一组数据 .

import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf
from sklearn import datasets
from tensorflow.python.framework import ops
ops.reset_default_graph()
sess = tf.Session()
iris = datasets.load_iris() # Load the data
x_vals = np.array([x[3] for x in iris.data])
y_vals = np.array([y[0] for y in iris.data])
batch_size = 25 # Declare batch size
x_data = tf.placeholder(shape=[None, 1], dtype=tf.float32) # Initialize placeholders
y_target = tf.placeholder(shape=[None, 1], dtype=tf.float32)
A = tf.Variable(tf.random_normal(shape=[1,1])) # Create variables
b = tf.Variable(tf.random_normal(shape=[1,1]))

# Declare model operations
model_output = tf.add(tf.matmul(x_data, A), b)

loss = tf.reduce_mean(tf.square(y_target - model_output)) # Declare loss function (L2 loss)
my_opt = tf.train.GradientDescentOptimizer(0.05) # Declare optimizer
train_step = my_opt.minimize(loss)
init = tf.global_variables_initializer() # Initialize variables
sess.run(init)

# Training loop
loss_vec = []
for i in range(100):
    rand_index = np.random.choice(len(x_vals), size=batch_size)
    rand_x = np.transpose([x_vals[rand_index]])
    rand_y = np.transpose([y_vals[rand_index]])
    sess.run(train_step, feed_dict={x_data: rand_x, y_target: rand_y})
    temp_loss = sess.run(loss, feed_dict={x_data: rand_x, y_target: rand_y})
    loss_vec.append(temp_loss)
    if (i+1)%25==0:
        print('Step #' + str(i+1) + ' A = ' + str(sess.run(A)) + ' b = ' + str(sess.run(b)))
        print('Loss = ' + str(temp_loss))

# Get the optimal coefficients获得最佳系数
[slope] = sess.run(A)
[y_intercept] = sess.run(b)

# Get best fit line获得最合适的线
best_fit = []
for i in x_vals:
  best_fit.append(slope*i+y_intercept)