Question

我对波士顿数据集做了线性回归问题，并获得了下一个结果：

损耗值不会随着数量增加而变化。这个错误的原因是什么？请帮我

import pandas as pd
import torch
import numpy as np
import torch.nn as nn
from sklearn import preprocessing
training_set=pd.read_csv('boston_data.csv')
training_set=training_set.to_numpy()
test_set=test_set.to_numpy()
inputs=training_set[:,0:13]
inputs=preprocessing.normalize(inputs)
target=training_set[:,13:14]
target=preprocessing.normalize(target)
inputs=torch.from_numpy(inputs)
target=torch.from_numpy(target)
test_set=torch.from_numpy(test_set)
w=torch.randn(13,1,requires_grad=True)
b=torch.randn(404,1,requires_grad=True)
def model(x):
    return x@w+b
pred=model(inputs.float())
def loss_MSE(x,y):
    ras=x-y
    return torch.sum(ras * ras) / ras.numel()
for i in range(100):
    pred=model(inputs.float())
    loss=loss_MSE(target,pred)
    loss.backward()
    with torch.no_grad():
        w -= w.grad * 1e-5
        b -= b.grad * 1e-5
        w.grad.zero_()
        b.grad.zero_()
    print(loss)

Answer 1

欢迎来到Stackoverflow

您的主循环很好（不过，您可以使自己的生活更轻松，但是，您应该阅读this），但是您的学习率（1e-5）很可能太低了。

我尝试了一个小的虚拟问题，很快以1e-2的学习速度解决了这个问题，而使用1e-5的学习时间则要长得多。虽然它确实会收敛，但是要经过100个以上的纪元。您提到您尝试增加纪元的数量，但没有写出实际进行多少次实验。请尝试增加此参数（学习率）以查看它是否可以解决您的问题。您也可以尝试通过<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.3.1/jquery.min.js"></script> <div class="header"> <div class="header__label">Grandtotal:<span id="total_grand">0</span> </div> <div class="header__label">XP:<span id="total_xp">0</span></div> </div> <div data-id='order'> <div class="order__header"> <div class="header__label">Order Total: <span class="order__total">000</span></div> <div class="header__label">Order Fee: <span class="order__fee">000</span></div> <div class="header__label">Order XP: <span class="xp__total--order">000</span></div> </div> <div data-id='history'> <div class="history__name">Item 1 (10xp)</div> <div><span class="history__sold">25</span>Sold</div> <div><span class="history__cost">10</span>Cost</div> <div><span class="history__subtotal">000</span>Subtotal</div> <div>Fee:<span class="history__fee">0</span></div> <div><span class="xp__total--history">000</span>XP</div> </div> <div data-id='history'> <div class="history__name">Item 2 (10xp)</div> <div><span class="history__sold">10</span>Sold</div> <div><span class="history__cost">50</span>Cost</div> <div><span class="history__subtotal">000</span>Subtotal</div> <div>Fee:<span class="history__fee">0</span></div> <div><span class="xp__total--history">000</span>XP</div> </div> </div> <div data-id='order'> <div class="order__header"> <div class="header__label">Order Total: <span class="order__total">000</span></div> <div class="header__label">Order Fee: <span class="order__fee">000</span></div> <div class="header__label">Order XP: <span class="xp__total--order">000</span></div> </div> <div data-id='history'> <div class="history__name">Item 1 (10xp)</div> <div><span class="history__sold">25</span>Sold</div> <div><span class="history__cost">10</span>Cost</div> <div><span class="history__subtotal">000</span>Subtotal</div> <div>Fee:<span class="history__fee">0</span></div> <div><span class="xp__total--history">000</span>XP</div> </div> <div data-id='history'> <div class="history__name">Item 2 (10xp)</div> <div><span class="history__sold">10</span>Sold</div> <div><span class="history__cost">50</span>Cost</div> <div><span class="history__subtotal">000</span>Subtotal</div> <div>Fee:<span class="history__fee">0</span></div> <div><span class="xp__total--history">000</span>XP</div> </div> </div>删除除法，这将具有相同的效果（除法也适用于渐变）。

下次，请提供一个small minimal example号，该号将无法运行，并有助于重现您的错误。这里的大部分代码都是数据加载，可以用两行虚拟数据生成代替。

线性回归中的损失值

1 个答案: