tensorflow梯度下降

import tensorflow as tf

# model parameters
w = tf.variable([0
], dtype=tf.float32)
#w初值及型別
b = tf.variable([0
], dtype=tf.float32)
#b初值及型別
# model input and output
x = tf.placeholder(tf.float32)
#x的型別
linear_model = w*x + b     
y = tf.placeholder(tf.float32)
#y的型別
# loss
loss = tf.reduce_sum(tf.square(linear_model - y)
)# sum of the squares
# optimizer
optimizer = tf.train.gradientdescentoptimizer(
0.001
)train = optimizer.minimize(loss)
# training data
x_train =[1
,2,3
,4,5
]y_train =[0
,-1,
-2,-
3,-4
]# training loop
init = tf.global_variables_initializer(
)sess = tf.session(
)sess.run(init)
# reset values to wrong
for i in
range
(1000):
#迴圈迭代次數
sess.run(train,
)# evaluate training accuracy
curr_w, curr_b, curr_loss = sess.run(
[w, b, loss],)
print
("w: %s b: %s loss: %s"
%(curr_w, curr_b, curr_loss)
)

例如：

import tensorflow as tf

# model parameters
w1 = tf.variable([0
], dtype=tf.float32) 
w2 = tf.variable([0
], dtype=tf.float32) 
b = tf.variable([0
], dtype=tf.float32)
# model input and output
x1 = tf.placeholder(tf.float32)
x2 = tf.placeholder(tf.float32)
linear_model = w1*x1 + w2*x2 + b
y = tf.placeholder(tf.float32)
# loss
loss = tf.reduce_sum(tf.square(linear_model - y)
)# sum of the squares
# optimizer
optimizer = tf.train.gradientdescentoptimizer(
0.001
)train = optimizer.minimize(loss)
# training data
x1_train =[1
,2,3
,4,5
]x2_train =[0
,2,1
,3,6
]y_train =[1
,4,4
,7,11
]# training loop
init = tf.global_variables_initializer(
)sess = tf.session(
)sess.run(init)
# reset values to wrong
for i in
range
(1000):
sess.run(train,
)# evaluate training accuracy
curr_w1,curr_w2, curr_b, curr_loss = sess.run(
[w1,w2, b, loss],)
print
("w1: %s w2: %s b: %s loss: %s"
%(curr_w1, curr_w2, curr_b, curr_loss)
)

可能受電腦配置影響，迭代次數多了之後會比較慢。

以開頭的**為例，引入matplotlib包，然後呼叫一些裡面的畫圖的函式。

import matplotlib.pyplot as plt  #引入matplotlib包

import tensorflow as tf
# model parameters
w = tf.variable([0
], dtype=tf.float32)
#w初值及型別
b = tf.variable([0
], dtype=tf.float32)
#b初值及型別
# model input and output
x = tf.placeholder(tf.float32)
#x型別
linear_model = w*x + b     
y = tf.placeholder(tf.float32)
#y型別
# loss
loss = tf.reduce_sum(tf.square(linear_model - y)
)# sum of the squares
# optimizer
optimizer = tf.train.gradientdescentoptimizer(
0.001
)train = optimizer.minimize(loss)
# training data
x_train =[1
,2,3
,4,5
]y_train =[0
,-1,
-2,-
3,-4
]# training loop
init = tf.global_variables_initializer(
)sess = tf.session(
)sess.run(init)
# reset values to wrong
for i in
range
(1000):
#迴圈迭代次數
sess.run(train,
)# evaluate training accuracy
curr_w, curr_b, curr_loss = sess.run(
[w, b, loss],)
print
("w: %s b: %s loss: %s"
%(curr_w, curr_b, curr_loss)
)line_values =
[curr_w * i + curr_b for i in x_train]
#根據已算出的斜率和截距寫線的關係式
plt.scatter(x_train,y_train,color=
'r')
#根據x，y的座標描點，紅色
plt.plot(x_train, line_values,
'b')
#畫線plt.show(
)#畫圖

TensorFlow中梯度下降函式

一 介紹 下面介紹在tensorflow中進行隨機梯度下降優化的函式。在tensorflow中通過乙個叫optimizer的優化器類進行訓練優化。二 梯度下降優化器 三 說明 在訓練過程中先例項化乙個優化函式如tf.train.gradientdescentoptimizer，並基於一定的學習率進行...

tensorflow實現梯度下降演算法

假設每次生成資料的公式為 每次計算出的損失函式為 這裡使用平方損失函式 每次採用梯度下降演算法，將梯度 設定為0.001，梯度下降公式為 將data初始化為二維的對應陣列 defgetloss w,b totalloss 0.0 radientw 0.0 radientb 0.0for i in r...

梯度下降 隨機梯度下降 批梯度下降

下面的h x 是要擬合的函式，j 損失函式，theta是引數，要迭代求解的值，theta求解出來了那最終要擬合的函式h 就出來了。其中m是訓練集的記錄條數，j是引數的個數。梯度下降法流程 1 先對 隨機賦值，可以是乙個全零的向量。2 改變 的值，使j 按梯度下降的方向減少。以上式為例 1 對於我們的...