def __init__(self,
train_file,
test_file,
prediction_file=None,
steps=30,
learn_rate=0.01,
delta=0.015,
factors=10,
init_mean=0.1,
init_stdev=0.1,
bias_learn_rate=0.005,
bias_reg=0.002,
random_seed=0):
MatrixFactorization.__init__(self,
train_file=train_file,
test_file=test_file,
prediction_file=prediction_file,
steps=steps,
learn_rate=learn_rate,
delta=delta,
factors=factors,
init_mean=init_mean,
init_stdev=init_stdev,
baseline=True,
bias_learn_rate=bias_learn_rate,
delta_bias=bias_reg,
random_seed=random_seed)
self.y = np.random.normal(self.init_mean, self.init_stdev,
(len(self.items), self.factors))
self.n_u = dict()
# |N(u)|^(-1/2)
for u, user in enumerate(self.users):
self.n_u[u] = np.power(
len(self.train_set["feedback"].get(user, [0])), -.5)
def get_mf(self):
MatrixFactorization(
train_file=self.train_file,
test_file=self.test_file,
output_file=self.output_file,
factors=int(self.factors),
learn_rate=self.learn_rate,
delta=self.delta,
sep=self.sep,
output_sep=self.output_sep).compute(verbose_evaluation=False)
def run_recommenders(self, r):
"""
1: Item KNN
2: User KNN
3: Matrix Factorization
4: SVD++
"""
flag = True
if not self.unlabeled_data[r]:
flag = False
return [], flag
else:
if r == 1:
rec = ItemKNN(self.labeled_files[r],
self.unlabeled_files[r],
as_similar_first=True)
rec.read_files()
rec.init_model()
rec.train_baselines()
rec.predict()
self.recommenders_predictions.setdefault(r, rec.predictions)
elif r == 2:
rec = UserKNN(self.labeled_files[r], self.unlabeled_files[r])
rec.read_files()
rec.init_model()
rec.train_baselines()
rec.predict()
self.recommenders_predictions.setdefault(r, rec.predictions)
elif r == 3:
rec = MatrixFactorization(self.labeled_files[r],
self.unlabeled_files[r],
random_seed=1,
baseline=True)
rec.read_files()
rec.init_model()
rec.fit()
rec.predict()
self.recommenders_predictions.setdefault(r, rec.predictions)
elif r == 4:
rec = SVDPlusPlus(self.labeled_files[r],
self.unlabeled_files[r],
random_seed=1)
rec.read_files()
rec.fit()
rec.predict()
self.recommenders_predictions.setdefault(r, rec.predictions)
else:
raise NameError('Invalid Recommender!')
return rec.predictions, flag
"""
from caserec.recommenders.rating_prediction.svdplusplus import SVDPlusPlus
from caserec.recommenders.rating_prediction.matrixfactorization import MatrixFactorization
from caserec.utils.cross_validation import CrossValidation
db = 'C:/Users/user/OneDrive/ml-100k/u.data'
folds_path = 'C:/Users/user/OneDrive/ml-100k/'
metadata_item = 'C:/Users/user/OneDrive/ml-100k/db_item_subject.dat'
sm_item = 'C:/Users/user/OneDrive/ml-100k/sim_item.dat'
metadata_user = '******'
sm_user = '******'
tr = 'C:/Users/user/OneDrive/ml-100k/folds/0/train.dat'
te = 'C:/Users/user/OneDrive/ml-100k/folds/0/test.dat'
"""
UserKNN
"""
# Cross Validation
recommender = MatrixFactorization()
# CrossValidation(input_file=db, recommender=recommender, dir_folds=folds_path, header=1, k_folds=5).compute()
# # Simple
MatrixFactorization(tr, te).compute()
SVDPlusPlus(tr, te).compute()
from caserec.utils.cross_validation import CrossValidation
db = '../../datasets/ml-100k/u.data'
folds_path = '../../datasets/ml-100k/'
metadata_item = '../../datasets/ml-100k/db_item_subject.dat'
sm_item = '../../datasets/ml-100k/sim_item.dat'
metadata_user = '******'
sm_user = '******'
tr = '../../datasets/ml-100k/folds/0/train.dat'
te = '../../datasets/ml-100k/folds/0/test.dat'
"""
UserKNN
"""
# Cross Validation
recommender = MatrixFactorization()
CrossValidation(input_file=db,
recommender=recommender,
dir_folds=folds_path,
header=1,
k_folds=5).compute()
# # Simple
# MatrixFactorization(tr, te).compute()
# SVDPlusPlus(tr, te).compute()
UserKNN
"""
# Cross Validation
# recommender = MatrixFactorization()
# CrossValidation(input_file=db, recommender=recommender, dir_folds=folds_path, header=1, k_folds=5).compute()
# # Simple
# MatrixFactorization(tr, te).compute()
# SVDPlusPlus(tr, te).compute()
train = "C:/dev/train.dat"
test = "C:/dev/test.dat"
metadata = "C:/dev/item_subject.dat"
# GSVDPlusPlus(train, test, metadata_file=metadata).compute()
similarity = "C:/dev/vsm.dat"
MatrixFactorization(train, test, random_seed=0, epochs=30,
stop_criteria=0.009).compute()
ItemMSMF(train,
test,
similarity_file=similarity,
neighbors=20,
random_seed=0,
baseline=True,
epochs=30,
stop_criteria=0.009).compute()