def test_on_batch(self, test_batch_data, test_batch_label, train_mode = False):
predictions = self.foward_pass(test_batch_data, train_mode = train_mode)
loss = np.mean(objective(self.loss).forward(predictions, test_batch_label))
acc = objective(self.loss).accuracy(predictions, test_batch_label)
return loss, acc
def train_on_batch(self, train_batch_data, train_batch_label):
predictions = self.foward_pass(train_batch_data, train_mode = True)
loss = np.mean(objective(self.loss).forward(predictions, train_batch_label))
acc = objective(self.loss).accuracy(predictions, train_batch_label)
self.backward_pass(objective(self.loss).backward(predictions, train_batch_label))
return loss, acc
def __init__(self,
epochs,
loss = 'binary_crossentropy',
init_method = 'he_normal',
optimizer = {},
penalty = 'lasso',
penalty_weight = 0,
l1_ratio = 0.5):
self.epochs = epochs
self.loss = objective(loss)
self.init_method = init(init_method)
self.optimizer = optimize(optimizer)
self.activate = activation('sigmoid')
self.regularization = regularize(penalty, penalty_weight, l1_ratio = l1_ratio)
def __init__(self,
epochs,
loss='mean_squared_error',
init_method='he_uniform',
optimizer={},
penalty='ridge',
penalty_weight=0.5,
l1_ratio=0.5):
self.epochs = epochs
self.loss = objective(loss)
self.init_method = init(init_method)
self.optimizer = optimize(optimizer)
self.regularization = regularize(penalty,
penalty_weight,
l1_ratio=l1_ratio)
input_data, input_label, test_size=0.3)
opt = register_opt(optimizer_name='sgd', momentum=0.01, learning_rate=0.001)
# model definition
model = LinearRegression(epochs=100,
optimizer=opt,
penalty='l1',
penalty_weight=0.8)
fit_stats = model.fit(train_data, train_label)
# fit_stats = model.fit_OLS(train_data, train_label) # ---- Ordinary Least Squares Method
targets = np.expand_dims(test_label, axis=1)
predictions = np.expand_dims(model.predict(test_data), axis=1)
mse = objective('mean_squared_error').forward(predictions, targets)
print('Mean Squared Error: {:.2f}'.format(mse))
model_name = 'boston_linear_regression'
plot_metric('accuracy_loss',
len(fit_stats['train_loss']),
fit_stats['train_acc'],
fit_stats['train_loss'],
model_name=model_name,
legend=['acc', 'loss'])
plot_regression_results(train_data,
train_label,
test_data,
test_label,
