def example():
"""simple test and performance measure
"""
num_user = 944
num_item = 1683
num_hidden = 200
iterations = 400
corruption = 0.2
train = build_ml_100k_train_binary3()
test = build_ml_100k_test_binary3()
train_matrix = build_user_item_matrix(num_user, num_item, train)
test_matrix = build_user_item_matrix(num_user, num_item, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
num_user = num_user - 1
num_item = num_item - 1
mf_model = CDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, corruption)
train = build_ml_100k_train_binary1()
test = build_ml_100k_test_binary1()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, corruption)
train = build_ml_100k_train_binary2()
test = build_ml_100k_test_binary2()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, corruption)
train = build_ml_100k_train_binary4()
test = build_ml_100k_test_binary4()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, corruption)
train = build_ml_100k_train_binary5()
test = build_ml_100k_test_binary5()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, corruption)
return mf_model
def example():
"""simple test and performance measure
"""
num_user = 6041
num_item = 3953
num_hidden = 500
iterations = 501
train = build_ml_1m_train_binary2()
test = build_ml_1m_test_binary2()
train_matrix = build_user_item_matrix(num_user, num_item, train)
test_matrix = build_user_item_matrix(num_user, num_item, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
num_user = num_user - 1
num_item = num_item - 1
mf_model = CDAE_ML_1m(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, 0.2)
train = build_ml_1m_train_binary4()
test = build_ml_1m_test_binary4()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_1m(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, 0.2)
train = build_ml_1m_train_binary5()
test = build_ml_1m_test_binary5()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_1m(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, 0.2)
train = build_ml_1m_train_binary3()
test = build_ml_1m_test_binary3()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_1m(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, 0.2)
train = build_ml_1m_train_binary1()
test = build_ml_1m_test_binary1()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = CDAE_ML_1m(train_matrix.shape[0], num_item, train_matrix,
test_matrix, num_hidden)
mf_model.estimate(iterations, 0.2)
return mf_model
def example():
"""simple test and performance measure
"""
num_user = 943
num_item = 1682
num_hidden = 200
iterations = 400
pre_iterations = 400
alpha = 0.4
imputation_ratio = 0.02
train = build_ml_100k_train_binary3()
test = build_ml_100k_test_binary3()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations, imputation_ratio)
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, output_matrix, num_hidden)
mf_model.estimate(iterations, alpha)
train = build_ml_100k_train_binary1()
test = build_ml_100k_test_binary1()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations, imputation_ratio)
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, output_matrix, num_hidden)
mf_model.estimate(iterations, alpha)
train = build_ml_100k_train_binary2()
test = build_ml_100k_test_binary2()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations, imputation_ratio)
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, output_matrix, num_hidden)
mf_model.estimate(iterations, alpha)
train = build_ml_100k_train_binary4()
test = build_ml_100k_test_binary4()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations, imputation_ratio)
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, output_matrix, num_hidden)
mf_model.estimate(iterations, alpha)
train = build_ml_100k_train_binary5()
test = build_ml_100k_test_binary5()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations, imputation_ratio)
mf_model = IDAE_ML_100k(train_matrix.shape[0], num_item, train_matrix,
test_matrix, output_matrix, num_hidden)
mf_model.estimate(iterations, alpha)
return mf_model
def example():
"""simple test and performance measure
"""
num_user = 944
num_item = 1683
num_hidden = 200
iterations = 400
pre_iterations = 400
train = build_ml_100k_train_binary3()
test = build_ml_100k_test_binary3()
train_matrix = build_user_item_matrix(num_user, num_item, train)
test_matrix = build_user_item_matrix(num_user, num_item, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
num_user = num_user - 1
num_item = num_item - 1
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations)
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
output_matrix, num_hidden)
mf_model.estimate(iterations)
train = build_ml_100k_train_binary1()
test = build_ml_100k_test_binary1()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations)
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
output_matrix, num_hidden)
mf_model.estimate(iterations)
train = build_ml_100k_train_binary2()
test = build_ml_100k_test_binary2()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations)
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
output_matrix, num_hidden)
mf_model.estimate(iterations)
train = build_ml_100k_train_binary4()
test = build_ml_100k_test_binary4()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations)
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
output_matrix, num_hidden)
mf_model.estimate(iterations)
train = build_ml_100k_train_binary5()
test = build_ml_100k_test_binary5()
train_matrix = build_user_item_matrix(num_user + 1, num_item + 1, train)
test_matrix = build_user_item_matrix(num_user + 1, num_item + 1, test)
train_matrix = train_matrix.todense()
test_matrix = test_matrix.todense()
train_matrix = train_matrix[1:, 1:]
test_matrix = test_matrix[1:, 1:]
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
train_matrix, num_hidden)
output_matrix = mf_model.preTrain(pre_iterations)
mf_model = DAE_with_imputation_ML_100k(train_matrix.shape[0], num_item,
train_matrix, test_matrix,
output_matrix, num_hidden)
mf_model.estimate(iterations)
return mf_model