def main_roman():
(features, success_labels, yard_labels, progress_labels) = extract_features(2009, 2014)
features, enc = encode_categorical_features(features, sparse=False)
print enc.vocabulary_
configurations = {'success': {'labels': success_labels, 'target': 'success', 'regression': False},
'yards': {'labels': yard_labels, 'target': 'yards', 'regression': True},
'progress': {'labels': progress_labels, 'target': 'progress', 'regression': True}}
selected_configuration = 'progress'
neural_network_prediction(features=features,
labels=configurations[selected_configuration]['labels'],
k=5,
team='all',
target_name=configurations[selected_configuration]['target'],
regression_task=configurations[selected_configuration]['regression'],
epochs = [10],
hidden_layers = [1, 10],
hidden_units = [10, 50, 100],
load_previous = True)
def main_brendan():
(features, labels, yards, progress) = extract_features(2009, 2014)
#compare_RBF_parameters(features, labels)
compute_regression_results(features,progress, "./regression_results_progress.txt")
compute_regression_results(features,yards, "./regression_results_yards.txt")
def main_valentin():
(features, labels, _, _) = extract_features(2014, 2014)
compare_RBF_parameters(features, labels)