5 - Linear Regression 구현

저희가 예전에 수학을 배울 때 저 선을 y = ax + b라고 했었는데

 

저희는 지금 이 선을 Wx + b라고 하겠습니다.

 

 

 

 

 

import tensorflow as tf
print(tf.__version__)

# 텐서플로우 버전 확인
# 만약 1.x가 나오면 아래의 코드 실행
# !pip uninstall tensorflow --yes
# !pip install tensorflow==2.0.0

import numpy as np
rng = np.random

# 파라메터 설정
learning_rate = 0.01
training_steps = 1000
display_step = 50

# 학습
X = np.array([1,2,3,4])
Y = np.array([1,2,3,4])
n_samples = X.shape[0]

# W b 설정
W = tf.Variable(rng.randn(), name="weight")
b = tf.Variable(rng.randn(), name="bias")

# Linear regression (Wx + b).
def linear_regression(x):
    return W * x + b

# 선에 주황색 줄들
def mean_square(y_pred, y_true):
    return tf.reduce_sum(tf.pow(y_pred-y_true, 2)) / (2 * n_samples)

# 경사하강법 적용
optimizer = tf.optimizers.SGD(learning_rate)

# optimizer 
def run_optimization():
    # Wrap computation inside a GradientTape for automatic differentiation.
    with tf.GradientTape() as g:
        pred = linear_regression(X)
        loss = mean_square(pred, Y)

    gradients = g.gradient(loss, [W, b])
    
    optimizer.apply_gradients(zip(gradients, [W, b]))

# 학습 gogo
for step in range(1, training_steps + 1):
    # Run the optimization to update W and b values.
    run_optimization()
    
    if step % display_step == 0:
        pred = linear_regression(X)
        loss = mean_square(pred, Y)
        print("step: %i, loss: %f, W: %f, b: %f" % (step, loss, W.numpy(), b.numpy()))
        
import matplotlib.pyplot as plt

# 시각화
plt.plot(X, Y, 'ro', label='Original data')
plt.plot(X, np.array(W * X + b), label='Fitted line')
plt.legend()
plt.show()

# 코드참조
# https://github.com/aymericdamien/TensorFlow-Examples/blob/master/tensorflow_v2/notebooks/2_BasicModels/linear_regression.ipynb

 

# 텐서플로우 버전 바꾸기
# !pip uninstall tensorflow --yes
# !pip install tensorflow==2.0.0

import tensorflow as tf
print(tf.__version__)

import numpy as np

x_train = [1,2,3,4]
y_train = [1.5,3,4.5,6]

model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Dense(1, input_dim = 1))

sgd = tf.keras.optimizers.SGD(lr=0.01)
model.compile(loss="mean_squared_error", optimizer=sgd)
model.fit(x_train, y_train, epochs=1000)

print(model.predict(np.array([4])))


# 코드참조
# https://www.youtube.com/watch?v=YT2Uy5x3gaA