深度学习之处理多维特征的输入

算法介绍

之前输入数据都是一维输入，一维输出，这里以这个数据为案例，输入8维的数据，这8维度分别代表8个特征，Y代表好坏情况

数学公式

代码

import numpy as np
import torch
import matplotlib.pyplot as plt

# prepare dataset
xy = np.loadtxt('PyTorch_exercise\diabetes.csv.gz', delimiter=',', dtype=np.float32)
x_data = torch.from_numpy(xy[:, :-1])  # 第一个‘：’是指读取所有行，第二个‘：’是指从第一列开始，最后一列不要
y_data = torch.from_numpy(xy[:, [-1]])  # [-1] 最后得到的是个矩阵


# design model using class
class Model(torch.nn.Module):
    def __init__(self):
        super(Model, self).__init__()
        self.linear1 = torch.nn.Linear(8, 6)  # 输入数据x的特征是8维，x有8个特征
        self.linear2 = torch.nn.Linear(6, 4)
        self.linear3 = torch.nn.Linear(4, 1)
        self.sigmoid = torch.nn.Sigmoid()  # 一个类

    def forward(self, x):
        x = self.sigmoid(self.linear1(x))
        x = self.sigmoid(self.linear2(x))
        x = self.sigmoid(self.linear3(x))  # y hat
        return x


model = Model()

# construct loss and optimizer
# criterion = torch.nn.BCELoss(size_average = True)
criterion = torch.nn.BCELoss(reduction='mean')
optimizer = torch.optim.SGD(model.parameters(), lr=0.1)

epoch_list = []
loss_list = []
# training cycle forward, backward, update
for epoch in range(100):
    y_pred = model(x_data)
    loss = criterion(y_pred, y_data)
    print(epoch, loss.item())
    epoch_list.append(epoch)
    loss_list.append(loss.item())

    optimizer.zero_grad()
    loss.backward()

    optimizer.step()

plt.plot(epoch_list, loss_list)
plt.ylabel('loss')
plt.xlabel('epoch')
plt.show()

结果分析

0 0.7374240756034851
1 0.7286046743392944
2 0.720624566078186
3 0.7134066224098206
4 0.7068801522254944
5 0.7009800672531128
6 0.6956475377082825
7 0.6908284425735474
8 0.6864737272262573
9 0.6825390458106995
10 0.6789836883544922
11 0.6757713556289673
12 0.6728686690330505
13 0.6702457070350647
14 0.6678754091262817
15 0.665733277797699
16 0.6637970209121704
17 0.6620467305183411
18 0.6604644656181335
19 0.6590337157249451
20 0.6577399373054504
21 0.6565698981285095
22 0.6555115580558777
23 0.6545541286468506
24 0.6536878347396851
25 0.6529038548469543
26 0.6521943211555481
27 0.6515522003173828
28 0.6509708166122437
29 0.6504444479942322
30 0.6499676704406738
31 0.6495358347892761
32 0.6491447687149048
33 0.6487904191017151
34 0.6484694480895996
35 0.6481783986091614
36 0.6479147672653198
37 0.647675633430481
38 0.6474589109420776
39 0.647262454032898
40 0.6470841765403748
41 0.6469224691390991
42 0.6467759013175964
43 0.6466427445411682
44 0.6465219855308533
45 0.6464123129844666
46 0.6463128328323364
47 0.6462224721908569
48 0.6461403965950012
49 0.6460658311843872
50 0.6459981203079224
51 0.6459364891052246
52 0.6458805203437805
53 0.6458296775817871
54 0.6457833647727966
55 0.6457412838935852
56 0.6457029581069946
57 0.6456679701805115
58 0.6456361413002014
59 0.6456071734428406
60 0.6455807685852051
61 0.6455566883087158
62 0.6455346345901489
63 0.6455145478248596
64 0.6454962491989136
65 0.6454795002937317
66 0.6454641222953796
67 0.6454500555992126
68 0.6454371809959412
69 0.6454253792762756
70 0.6454145312309265
71 0.645404577255249
72 0.6453953385353088
73 0.6453868746757507
74 0.6453790664672852
75 0.6453719735145569
76 0.6453652381896973
77 0.6453590393066406
78 0.6453533172607422
79 0.6453480124473572
80 0.6453430652618408
81 0.6453384757041931
82 0.6453341245651245
83 0.6453300714492798
84 0.6453263163566589
85 0.6453227400779724
86 0.6453193426132202
87 0.6453161239624023
88 0.6453131437301636
89 0.6453103423118591
90 0.6453076004981995
91 0.6453049778938293
92 0.6453025341033936
93 0.6453002095222473
94 0.6452978849411011
95 0.6452957391738892
96 0.6452935934066772
97 0.6452915072441101
98 0.6452895402908325
99 0.6452876925468445

因本人学的不是很深，所以有的可能没有写清楚，还请见谅