reduced_fit_times = {'PCA': [], "RCA": []}
reduced_predict_times = {'PCA': [], "RCA": []}
# transform stuff, but don't transform the ownership of this file, which is Boyko Todorov's
for a in range(num_iter):
x_train, x_test, y_train, y_test = data_service. \
load_and_split_data(scale_data=scale_data, transform_data=transform_data, random_slice=random_slice,
random_seed=random_seed, dataset=dataset, test_size=test_size)
if dataset == 'kdd':
y_train, y_test = convert_to_binary_service.convert(y_train, y_test, 11)
nn_learner = NNLearner(hidden_layer_sizes=nn_hidden_layer_sizes, max_iter=200, solver=nn_solver,
activation=nn_activation,
alpha=alpha, learning_rate=nn_learning_rate, learning_rate_init=nn_learning_rate_init)
nn_accuracy_score, nn_fit_time, nn_predict_time = nn_learner.fit_predict_score(x_train.copy(), y_train.copy(),
x_test.copy(), y_test.copy())
original_accuracies.append(nn_accuracy_score)
original_non_reduced_fit_times.append(nn_fit_time)
original_non_reduced_predict_times.append(nn_predict_time)
print("Iter {0}. Orig score: {1}, fit_time: {2}, predict_time: {3}".format(a, nn_accuracy_score, nn_fit_time, nn_predict_time))
for reduction_algo in reduction_algos:
nn_learner = NNLearner(hidden_layer_sizes=nn_hidden_layer_sizes, max_iter=200, solver=nn_solver,
activation=nn_activation,
alpha=alpha, learning_rate=nn_learning_rate,
dataset = 'breast_cancer'
test_size = 0.5
nn_activation = 'relu'
alpha = 0.0001
nn_hidden_layer_sizes = (10, )
nn_learning_rate = 'constant'
nn_learning_rate_init = 0.01
nn_solver = 'lbfgs'
#{'activation': 'relu', 'alpha': 0.0001, 'hidden_layer_sizes': (100,), 'learning_rate_init': 0.01, 'solver': 'lbfgs'}
nn_learner = NNLearner(hidden_layer_sizes=nn_hidden_layer_sizes,
max_iter=200,
solver=nn_solver,
activation=nn_activation,
alpha=alpha,
learning_rate=nn_learning_rate,
learning_rate_init=nn_learning_rate_init)
x_train, x_test, y_train, y_test = data_service.load_and_split_data(
scale_data=scale_data,
transform_data=transform_data,
random_slice=random_slice,
random_seed=random_seed,
dataset=dataset,
test_size=test_size)
nn_accuracy_score, nn_fit_time, nn_predict_time = nn_learner.fit_predict_score(
x_train, y_train, x_test, y_test)
from sklearn.multiclass import OneVsRestClassifier
dt_learner = DTLearner()
dt_learnerOnevsRest = OneVsRestClassifier(dt_learner.estimator)
#--------------------------------
nn_hidden_layer_sizes = (100, )
nn_solver = 'lbfgs'
nn_activation = 'relu'
alpha = 0.0001 # regularization term coefficient
nn_learning_rate = 'constant'
nn_learning_rate_init = 0.0001
nn_learner = NNLearner(hidden_layer_sizes=nn_hidden_layer_sizes,
max_iter=200,
solver=nn_solver,
activation=nn_activation,
alpha=alpha,
learning_rate=nn_learning_rate,
learning_rate_init=nn_learning_rate_init)
nn_learner_non_scaled = NNLearner(hidden_layer_sizes=nn_hidden_layer_sizes,
max_iter=200,
solver=nn_solver,
activation=nn_activation,
alpha=alpha,
learning_rate=nn_learning_rate,
learning_rate_init=nn_learning_rate_init)
#-------------------------------
n_neighbors = 5
weights = 'distance'
algorithm = 'auto'