第三讲代码部分实现,根据错错莫整理
1、梯度下降算法
import matplotlib.pyplot as plt
x_data = [1.0, 2.0, 3.0]
y_data = [2.0, 4.0, 6.0]
#预猜测w为1.0
w = 1.0
def forward(x):
return x * w
def cost(xs, ys):
cost = 0
for x, y in zip(xs, ys):
# 计算y_hat
y_pred = forward(x)
# 计算平方和(y_hat - y)^2
cost += (y-y_pred) ** 2
# 求均值
return cost / len(xs)
# 计算梯度
def gradients(xs, ys):
grad = 0
for x, y in zip(xs, ys):
grad += 2 * x * (x*w - y)
return grad / len(xs)
print('Predict (before training)', 4, forward(4))
# 进行100轮的训练
epoch_list = []
cost_list = []
for epoch in range(100):
cost_val = cost(x_data, y_data)
grad_val = gradients(x_data, y_data)
# w=w-学习率*梯度
w -= 0.01 * grad_val
print('Epoch:', epoch, 'w=', w, 'loss=', cost_val)
epoch_list.append(epoch)
cost_list.append(cost_val)
print('Predict (after training)', 4, forward(4))
plt.plot(epoch_list, cost_list)
plt.ylabel('Cost')
plt.xlabel('epoch')
plt.show()
效果图:
两者的不同:
随机梯度下降法和梯度下降法的主要区别在于:
- 损失函数由cost()更改为loss()。cost是计算所有训练数据的损失,loss是计算一个训练函数的损失。
- 梯度函数gradient()由计算所有训练数据的梯度更改为计算一个训练数据的梯度。
- 本算法中的随机梯度主要是指,每次拿一个训练数据来训练,然后更新梯度参数。本算法中梯度总共更新100(epoch)x3 =
300次。梯度下降法中梯度总共更新100(epoch)次。
2、随机梯度下降算法
import matplotlib.pyplot as plt
x_data = [1.0, 2.0, 3.0]
y_data = [2.0, 4.0, 6.0]
#预猜测w为1.0
w = 1.0
def forward(x):
return x * w
def loss(x, y):
# 计算y_hat
y_pred = forward(x)
return (y_pred - y) ** 2
# 计算梯度
def gradients(x, y):
return 2*x*(x*w - y)
print('Predict (before training)', 4, forward(4))
# 进行100轮的训练
# Update weight by every grad of sample of train set
epoch_list = []
loss_list = []
for epoch in range(100):
for x, y in zip(x_data, y_data):
grad = gradients(x, y)
# w=w-学习率*梯度
w -= 0.01 * grad
print("\t grad:", x, y, grad)
#计算损失
l = loss(x, y)
print("progress:", epoch, "w=", w, "loss=", l)
epoch_list.append(epoch)
loss_list.append(l)
print('Predict (after training)', 4, forward(4))
plt.plot(epoch_list, loss_list)
plt.ylabel('loss')
plt.xlabel('epoch')
plt.show()
效果图: