input=activ_pool1,
Kernels=Kernels_2,
stride=stride_2,
padding=0,
non_linearialty='ReLu')
activ_pool2 = functions.poollayer(input=out_a_pool2,
type_pool='max',
pool_size=pool_size_2)
z_pool2 = functions.poollayer(input=out_z_pool2,
type_pool='max',
pool_size=pool_size_2)
zs, logits = functions.mlp(input=np.ravel(activ_pool2),
weights=weights,
biases=biases,
num_hidden=num_hidden,
sizes=sizes,
non_linearialty='sigmoid',
output_size=output_size)
logits[-1] = functions.softmax(zs[-1])
delta = backprop.cost_derivative(
logits[-1], y_train) * backprop.softmax_grad(zs[-1])
der_beta = []
for i in range(sizes[0]):
der_beta.append(delta * logits[-2][i])
der_beta = np.transpose(np.array(der_beta))
q = 0
for i in range(10):
print('KNN')
k_values = [1, 2, 5, 7, 10]
functions.knn(X_train, X_test, Y_train, Y_test, k_values, True, ['uniform', 'distance'],
cfg.default.student_figures, 'knn')
print('Decission Tree Regression')
max_depths = [1, 10, 30, 50, 100, 300]
min_weight_fraction_leafs = [.0, .125, .25, .375, .5]
min_samples_leaf=[1, 10, 100, 200]
functions.decision_tree(X_train, X_test, Y_train, Y_test, max_depths, min_weight_fraction_leafs, min_samples_leaf,
cfg.default.student_figures, 'dtree')
print('MLP')
scaler = preprocessing.StandardScaler().fit(X_train)
X_train_scaled = scaler.transform(X_train)
X_test_scaled = scaler.transform(X_test)
max_iteration = 1000
solver = 'lbfgs' # lbfgs, adam, sgd
alpha = [0.001, 0.0001, 0.00001]
list_hidden_layer_sizes = [[10], [5, 2, 5], [60, 20]]
functions.mlp(X_train_scaled, X_test_scaled, Y_train, Y_test, max_iteration, solver, alpha, list_hidden_layer_sizes,
cfg.default.student_figures, 'mlp')
k_values = [1, 2, 5, 7, 10]
functions.knn(X_train, X_test, y_train, y_test, k_values, True,
['uniform', 'distance'], cfg.default.real_estate_figures, 'knn')
print('Decission Tree Regression')
max_depths = [1, 10, 30, 50, 100, 300]
min_weight_fraction_leafs = [.0, .125, .25, .375, .5]
min_samples_leaf = [1, 10, 100, 200]
functions.decision_tree(X_train, X_test, y_train, y_test, max_depths,
min_weight_fraction_leafs, min_samples_leaf,
cfg.default.real_estate_figures, 'dtree')
print('MLP')
scaler = preprocessing.StandardScaler().fit(X_train)
X_train_scaled = scaler.transform(X_train)
X_test_scaled = scaler.transform(X_test)
max_iteration = 1000
solver = 'lbfgs' # lbfgs, adam, sgd
alpha = [0.001, 0.0001, 0.00001]
list_hidden_layer_sizes = [[10], [5, 2, 5], [60, 20]]
functions.mlp(X_train_scaled, X_test_scaled, y_train, y_test, max_iteration,
solver, alpha, list_hidden_layer_sizes,
cfg.default.real_estate_figures, 'mlp')