def training_step(self, train_batch, batch_idx):
"""
Training the data as batches and returns training loss on each batch
:param train_batch: Batch data
:param batch_idx: Batch indices
:return: output - Training loss
"""
output = {}
x, y = train_batch
prob_mask = self.forward(x)
loss = self.criterion(prob_mask, y.type(torch.long))
acc = accuracy(torch.argmax(prob_mask, dim=1).float(), y)
output['acc'] = torch.tensor(acc)
iter_iou, iter_count = iou_fnc(
torch.argmax(prob_mask, dim=1).float(), y, self.args['n_class'])
for i in range(self.args['n_class']):
output['iou_' + str(i)] = torch.tensor(iter_iou[i])
output['iou_cnt_' + str(i)] = torch.tensor(iter_count[i])
output['loss'] = loss
return output
def validation_step(self, test_batch, batch_idx):
"""
Predicts on the test dataset to compute the current performance of the model.
:param test_batch: Batch data
:param batch_idx: Batch indices
:return: output - Validation performance
"""
output = {}
x, y = test_batch
prob_mask = self.forward(x)
loss = self.criterion(prob_mask, y.type(torch.long))
acc = accuracy(torch.argmax(prob_mask, dim=1).float(), y)
output['val_acc'] = torch.tensor(acc)
iter_iou, iter_count = iou_fnc(
torch.argmax(prob_mask, dim=1).float(), y, self.args['n_class'])
for i in range(self.args['n_class']):
output['val_iou_' + str(i)] = torch.tensor(iter_iou[i])
output['val_iou_cnt_' + str(i)] = torch.tensor(iter_count[i])
output['val_loss'] = loss
return output
def classification():
# Generate a random binary classification problem.
X, y = make_classification(
n_samples=1000,
n_features=20,
n_informative=10,
random_state=1111,
n_classes=2,
class_sep=2.5,
)
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.1,
random_state=1111)
for s in ['svd', 'eigen']:
p = PCA(10, solver=s)
# fit PCA with training data, not entire dataset
p.fit(X_train)
X_train_reduced = p.transform(X_train)
X_test_reduced = p.transform(X_test)
model = LogisticRegression(lr=0.01, max_iters=10, penalty='l2', C=0.01)
model.fit(X_train_reduced, y_train)
predictions = model.predict(X_test_reduced)
print('classification accuracy', accuracy(y_test, predictions))
def classificationfm():
# Generate a random binary classification problem.
X, y = make_classification(
n_samples=100,
n_features=20,
n_informative=10,
random_state=1111,
n_classes=2,
class_sep=2.5,
)
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.1,
random_state=1111)
model = FMClassifier(n_components=2, max_iter=20)
model.fit(X_train, y_train)
predictions = model.predict(X_test)
print('classification accuracy', accuracy(y_test, predictions))