/
ehr_model_finetuning.py
145 lines (104 loc) · 5.81 KB
/
ehr_model_finetuning.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
from models.ehr_classifier import EHR_classifier
from preprocessing import ehr_preprocessing
from dataset.EhrDataset import EhrDataset
import torch
import torch.nn as nn
from tqdm import tqdm
from models import model_checkpoints
from models.ehr_MLP import EHR_MLP
from visualization import plot_metrics
torch.manual_seed(0)
device = "cpu" if torch.cuda.is_available() else "cuda"
#loading data
X_ehr_train_processed, y_train = ehr_preprocessing.load_data('/home/santosh.sanjeev/PE-Research/RadFusion-Dataset/dataset/multimodalpulmonaryembolismdataset/new_data_csv/vision&emr_train_ed.csv')
X_ehr_val_processed, y_val = ehr_preprocessing.load_data('/home/santosh.sanjeev/PE-Research/RadFusion-Dataset/dataset/multimodalpulmonaryembolismdataset/new_data_csv/vision&emr_val_ed.csv')
X_ehr_test_processed, y_test = ehr_preprocessing.load_data('/home/santosh.sanjeev/PE-Research/RadFusion-Dataset/dataset/multimodalpulmonaryembolismdataset/new_data_csv/vision&emr_test_ed.csv')
X_train_ehr, X_val_ehr, X_test_ehr = ehr_preprocessing.feature_selection(X_ehr_train_processed, y_train, X_ehr_val_processed, X_ehr_test_processed)
# print('hiii', X_train_ehr.shape, type(X_train_ehr), X_train_ehr.dtype)
# print('hellloo', X_train_img.dtype, X_train_img.shape)
X_train_ehr, X_val_ehr, X_test_ehr = ehr_preprocessing.normalize(X_train_ehr, X_val_ehr, X_test_ehr)
print(X_train_ehr.dtype, X_train_ehr.shape, X_val_ehr.dtype, X_val_ehr.shape, X_test_ehr.dtype, X_test_ehr.shape)
#tsne-plots
plot_metrics.tsne_plot(y_train, path = './visualization/tsne_plots/ehr', name = 'ehr_train_data.png', img_data = X_train_ehr)
plot_metrics.tsne_plot(y_val, path = './visualization/tsne_plots/ehr', name = 'ehr_val_data.png', img_data = X_val_ehr)
plot_metrics.tsne_plot(y_test, path = './visualization/tsne_plots/ehr', name = 'ehr_test_data.png', img_data = X_test_ehr)
train_set = EhrDataset(X_train_ehr, y_train)
val_set = EhrDataset(X_val_ehr, y_val)
test_set = EhrDataset(X_test_ehr, y_test)
#dataloaders
train_loader = torch.utils.data.DataLoader(train_set, batch_size=16, shuffle=True)
val_loader = torch.utils.data.DataLoader(val_set, batch_size=16, shuffle=False)
test_loader = torch.utils.data.DataLoader(test_set, batch_size=16, shuffle=False)
pretrained_model = EHR_MLP()
ehr_model = model_checkpoints.load_ckp('./checkpoints/pretrained_ehr_model.pt', pretrained_model)
#intialising models
ehr_model = ehr_model.to(device)
ehr_classifier = EHR_classifier(ehr_model)
ehr_classifier = ehr_classifier.to(device)
# specify loss function
criterion = nn.CrossEntropyLoss()
# specify optimizer
optimizer = torch.optim.Adam(ehr_classifier.parameters(), lr=0.01)
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=20, gamma=0.1)
epochs= 80
valid_acc_max = 0
epochs_list=[]
training_loss=[]
validation_loss=[]
training_acc = []
validation_acc = []
for epoch in range(0, epochs):
train_loss=0
test_loss=0
train_total=0
train_correct=0
epochs_list.append(epoch)
print("epoch number: {0}".format(epoch))
ehr_classifier.train()
with tqdm(train_loader, unit = 'batch') as tepoch:
for batch_idx, (ehr_train_data, train_labels) in enumerate(tepoch):
ehr_train_data = ehr_train_data.to(device)
train_labels = train_labels.to(device)
output = ehr_classifier(ehr_train_data.float())
optimizer.zero_grad()
loss = criterion(output,train_labels)
loss.backward()
optimizer.step()
train_loss += loss.item()
_, predicted = output.max(1)
train_total += train_labels.size(0)
train_correct += predicted.eq(train_labels).sum().item()
with torch.no_grad():
test_total=0
test_correct= 0
ehr_classifier.eval()
with tqdm(val_loader, unit ="batch") as tepoch:
for batch_idx ,(ehr_val_data, val_labels) in enumerate(tepoch):
ehr_val_data = ehr_val_data.to(device)
val_labels = val_labels.to(device)
y_pred_test = ehr_classifier(ehr_val_data.float())
loss = criterion(y_pred_test, val_labels)
test_loss+=loss.item()
_, predicted = y_pred_test.max(1)
test_total += val_labels.size(0)
test_correct += predicted.eq(val_labels).sum().item()
ehr_checkpoint = {
'epoch': epoch + 1,
'valid_acc_max': 100.*test_correct/test_total,
'state_dict': ehr_classifier.state_dict(),
'optimizer': optimizer.state_dict(),
}
if 100.*test_correct/test_total > valid_acc_max:
# save checkpoint as best model
model_checkpoints.save_ckp(ehr_checkpoint, True, './checkpoints/ehr_classifier.pt')
valid_acc_max = 100.*test_correct/test_total
training_loss.append(train_loss/(batch_idx+1))
validation_loss.append(test_loss/(batch_idx+1))
training_acc.append(100.*train_correct/train_total)
validation_acc.append(100.*test_correct/test_total)
print('test loss: {:.4f} train_accuracy: {:.4f} test_accuracy: {:.4f} valid_acc{:.4f}'.format(test_loss/(batch_idx+1), 100.*train_correct/train_total, 100.*test_correct/test_total, valid_acc_max))
scheduler.step()
plot_metrics.plot_loss(epochs_list, training_loss,'Training Loss','./visualization/classifier/ehr_classifier/loss','ehr_classifier_training_loss.png')
plot_metrics.plot_loss(epochs_list, validation_loss,'Validation Loss' ,'./visualization/classifier/ehr_classifier/loss','ehr_classifier_validation_loss.png')
plot_metrics.plot_acc(epochs_list, training_acc,'Training Loss','./visualization/classifier/ehr_classifier/acc','ehr_classifier_training_acc.png')
plot_metrics.plot_acc(epochs_list, validation_acc,'Validation Loss' ,'./visualization/classifier/ehr_classifier/acc','ehr_classifier_validation_acc.png')