機器學習筆記(2)：線性迴歸-使用gluon

程式碼來自：https://zh.gluon.ai/chapter_supervised-learning/linear-regression-gluon.html

from mxnet import ndarray as nd
from mxnet import autograd
from mxnet import gluon

num_inputs = 2
num_examples = 1000

true_w = [2, -3.4]
true_b = 4.2

X = nd.random_normal(shape=(num_examples, num_inputs)) #1000行，2列的資料集
y = true_w[0] * X[:, 0] + true_w[1] * X[:, 1] + true_b #已知答案的結果
y += .01 * nd.random_normal(shape=y.shape) #加入噪音

#1 隨機讀取10行資料
batch_size = 10
dataset = gluon.data.ArrayDataset(X, y)
data_iter = gluon.data.DataLoader(dataset, batch_size, shuffle=True)

#2 定義迴歸模型
net = gluon.nn.Sequential()
net.add(gluon.nn.Dense(1))

#3 引數初始化
net.initialize()

#4 損失函式
square_loss = gluon.loss.L2Loss()

#5 指定訓練方法
trainer = gluon.Trainer(net.collect_params(), 'sgd', {'learning_rate': 0.1})
    
#6 訓練
epochs = 5
batch_size = 10
for e in range(epochs):
    total_loss = 0
    for data, label in data_iter:
        with autograd.record():
            output = net(data)
            loss = square_loss(output, label)
        loss.backward()
        trainer.step(batch_size)
        total_loss += nd.sum(loss).asscalar()
    print("Epoch %d, average loss: %f" % (e, total_loss/num_examples))

#7 輸出結果
dense = net[0]
print(true_w)
print(dense.weight.data())
print(true_b)
print(dense.bias.data())

相對上一篇純手動的處理方式，用gluon後程式碼明顯更精簡了。