[머신러닝-Tensorflow]Lec-07 multi-variable linear regression을 TensorFlow에서 구현하기

1. Hypothesis None using matrix -1  

import tensorflow as tf
	tf.set_random_seed(777) # for reproducibility
	x1_data = [73., 93., 89., 96., 73.]
	x2_data = [80., 88., 91., 98., 66.]
	x3_data = [75., 93., 90., 100., 70.]
	y_data = [152., 185., 180., 196., 142.]
	# placeholders for a tensor that will be always fed.
	x1 = tf.placeholder(tf.float32)
	x2 = tf.placeholder(tf.float32)
	x3 = tf.placeholder(tf.float32)
	Y = tf.placeholder(tf.float32)
	w1 = tf.Variable(tf.random_normal([1]), name='weight1')
	w2 = tf.Variable(tf.random_normal([1]), name='weight2')
	w3 = tf.Variable(tf.random_normal([1]), name='weight3')
	b = tf.Variable(tf.random_normal([1]), name='bias')
	hypothesis = x1 * w1 + x2 * w2 + x3 * w3 + b
	print(hypothesis)
	# cost/loss function
	cost = tf.reduce_mean(tf.square(hypothesis - Y))
	# Minimize. Need a very small learning rate for this data set
	optimizer = tf.train.GradientDescentOptimizer(learning_rate=1e-5)
	train = optimizer.minimize(cost)
	# Launch the graph in a session.
	sess = tf.Session()
	# Initializes global variables in the graph.
	sess.run(tf.global_variables_initializer())
	for step in range(2001):
	cost_val, hy_val, _ = sess.run([cost, hypothesis, train],
	feed_dict={x1: x1_data, x2: x2_data, x3: x3_data, Y: y_data})
	if step % 10 == 0:
	print(step, "Cost: ", cost_val, "\nPrediction:\n", hy_val)

2. Hypothesis using matrix -1

import tensorflow as tf
	tf.set_random_seed(777) # for reproducibility
	x_data = [[73., 80., 75.],
	[93., 88., 93.],
	[89., 91., 90.],
	[96., 98., 100.],
	[73., 66., 70.]]
	y_data = [[152.],
	[185.],
	[180.],
	[196.],
	[142.]]
	# placeholders for a tensor that will be always fed.
	X = tf.placeholder(tf.float32, shape=[None, 3])
	Y = tf.placeholder(tf.float32, shape=[None, 1]) # none의 의미는 내가 원하는 개수만큼 처리할수 있다는 의미이고, X의 3은 인스턴스의 개수가 3개라는 의미 Y에서의 1은 인스턴스의 개수가 1개
	W = tf.Variable(tf.random_normal([3, 1]), name='weight')
	b = tf.Variable(tf.random_normal([1]), name='bias')
	# Hypothesis
	hypothesis = tf.matmul(X, W) + b
	# Simplified cost/loss function
	cost = tf.reduce_mean(tf.square(hypothesis - Y))
	# Minimize
	optimizer = tf.train.GradientDescentOptimizer(learning_rate=1e-5)
	train = optimizer.minimize(cost)
	# Launch the graph in a session.
	sess = tf.Session()
	# Initializes global variables in the graph.
	sess.run(tf.global_variables_initializer())
	for step in range(2001):
	cost_val, hy_val, _ = sess.run(
	[cost, hypothesis, train], feed_dict={X: x_data, Y: y_data})
	if step % 10 == 0:
	print(step, "Cost: ", cost_val, "\nPrediction:\n", hy_val)

3. 출력 결과

(0, 'Cost: ', 752.86066, '\nPrediction:\n', array([[133.76956],

       [151.47867],

       [154.42642],

       [165.63803],

       [114.93678]], dtype=float32))

(10, 'Cost: ', 17.752771, '\nPrediction:\n', array([[157.73633],

       [180.31572],

       [182.82404],

       [196.56433],

       [136.93816]], dtype=float32))

(20, 'Cost: ', 17.652256, '\nPrediction:\n', array([[157.79259],

       [180.41405],

       [182.90495],

       [196.65446],

       [137.01912]], dtype=float32))

(30, 'Cost: ', 17.558985, '\nPrediction:\n', array([[157.77658],

       [180.42542],

       [182.9002 ],

       [196.65134],

       [137.03374]], dtype=float32))

(40, 'Cost: ', 17.466183, '\nPrediction:\n', array([[157.76039],

       [180.4365 ],

       [182.8952 ],

       [196.64792],

       [137.04811]], dtype=float32))

(50, 'Cost: ', 17.373915, '\nPrediction:\n', array([[157.74423],

       [180.44754],

       [182.89023],

       [196.64452],

       [137.06242]], dtype=float32))

(60, 'Cost: ', 17.28209, '\nPrediction:\n', array([[157.72815],

       [180.45859],

       [182.88528],

       [196.64116],

       [137.07675]], dtype=float32))

(70, 'Cost: ', 17.190739, '\nPrediction:\n', array([[157.71205],

       [180.46956],

       [182.88033],

       [196.63776],

       [137.091  ]], dtype=float32))

(80, 'Cost: ', 17.09996, '\nPrediction:\n', array([[157.69604],

       [180.48053],

       [182.87541],

       [196.63438],

       [137.10522]], dtype=float32))

(90, 'Cost: ', 17.00963, '\nPrediction:\n', array([[157.68007],

       [180.49147],

       [182.87048],

       [196.63103],

       [137.1194 ]], dtype=float32))

(100, 'Cost: ', 16.919806, '\nPrediction:\n', array([[157.66414],

       [180.50237],

       [182.86557],

       [196.62767],

       [137.13354]], dtype=float32))

(110, 'Cost: ', 16.830479, '\nPrediction:\n', array([[157.64824],

       [180.51323],

       [182.86069],

       [196.62433],

       [137.14764]], dtype=float32))

(120, 'Cost: ', 16.741568, '\nPrediction:\n', array([[157.63239],

       [180.52408],

       [182.85579],

       [196.621  ],

       [137.16171]], dtype=float32))

(130, 'Cost: ', 16.65318, '\nPrediction:\n', array([[157.61658],

       [180.5349 ],

       [182.85092],

       [196.61768],

       [137.17574]], dtype=float32))

(140, 'Cost: ', 16.565327, '\nPrediction:\n', array([[157.60083],