# Copyright 2018 The Cornac Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
import numpy as np
from ..recommender import Recommender
from ..recommender import ANNMixin, MEASURE_DOT
[docs]
class PCRL(Recommender, ANNMixin):
"""Probabilistic Collaborative Representation Learning.
Parameters
----------
k: int, optional, default: 100
The dimension of the latent factors.
z_dims: Numpy 1d array, optional, default: [300]
The dimensions of the hidden intermdiate layers 'z' in the order \
[dim(z_L), ...,dim(z_1)], please refer to Figure 1 in the orginal paper for more details.
max_iter: int, optional, default: 300
Maximum number of iterations (number of epochs) for variational PCRL.
batch_size: int, optional, default: 300
The batch size for SGD.
learning_rate: float, optional, default: 0.001
The learning rate for SGD.
aux_info: see "cornac/examples/pcrl_example.py" in the GitHub repo for an example of how to use \
cornac's graph modality provide item auxiliary data (e.g., context, text, etc.) for PCRL.
name: string, optional, default: 'PCRL'
The name of the recommender model.
trainable: boolean, optional, default: True
When False, the model is not trained and Cornac assumes that the model already \
pre-trained (Theta, Beta and Xi are not None).
w_determinist: boolean, optional, default: True
When True, determinist wheights "W" are used for the generator network, \
otherwise "W" is stochastic as in the original paper.
init_params: dictionary, optional, default: None
List of initial parameters, e.g., init_params = {'G_s':G_s, 'G_r':G_r, 'L_s':L_s, 'L_r':L_r}.
Theta: ndarray, shape (n_users, k)
The expected user latent factors.
Beta: ndarray, shape (n_items, k)
The expected item latent factors.
G_s: ndarray, shape (n_users, k)
Represent the "shape" parameters of Gamma distribution over Theta.
G_r: ndarray, shape (n_users, k)
Represent the "rate" parameters of Gamma distribution over Theta.
L_s: ndarray, shape (n_items, k)
Represent the "shape" parameters of Gamma distribution over Beta.
L_r: ndarray, shape (n_items, k)
Represent the "rate" parameters of Gamma distribution over Beta.
References
----------
* Salah, Aghiles, and Hady W. Lauw. Probabilistic Collaborative Representation Learning for Personalized Item Recommendation. \
In UAI 2018.
"""
def __init__(
self,
k=100,
z_dims=[300],
max_iter=300,
batch_size=300,
learning_rate=0.001,
name="PCRL",
trainable=True,
verbose=False,
w_determinist=True,
init_params=None,
):
Recommender.__init__(self, name=name, trainable=trainable, verbose=verbose)
self.k = k
self.z_dims = z_dims # the dimension of the second hidden layer (we consider a 2-layers PCRL)
self.max_iter = max_iter
self.batch_size = batch_size
self.learning_rate = learning_rate
self.w_determinist = w_determinist
# Init params if provided
self.init_params = {} if init_params is None else init_params
self.Theta = self.init_params.get("Theta", None)
self.Beta = self.init_params.get("Beta", None)
self.Gs = self.init_params.get("G_s", None)
self.Gr = self.init_params.get("G_r", None)
self.Ls = self.init_params.get("L_s", None)
self.Lr = self.init_params.get("L_r", None)
[docs]
def fit(self, train_set, val_set=None):
"""Fit the model to observations.
Parameters
----------
train_set: :obj:`cornac.data.Dataset`, required
User-Item preference data as well as additional modalities.
val_set: :obj:`cornac.data.Dataset`, optional, default: None
User-Item preference data for model selection purposes (e.g., early stopping).
Returns
-------
self : object
"""
Recommender.fit(self, train_set, val_set)
# X = sp.csc_matrix(self.train_set.matrix)
if self.trainable:
from .pcrl import PCRL_
# use pre-trained params if exists, otherwise from constructor
init_params = {
"G_s": self.Gs,
"G_r": self.Gr,
"L_s": self.Ls,
"L_r": self.Lr,
}
# instanciate pcrl
# train_aux_info = train_set.item_graph.matrix[:self.train_set.num_items, :self.train_set.num_items]
pcrl_ = PCRL_(
train_set=train_set,
k=self.k,
z_dims=self.z_dims,
n_epoch=self.max_iter,
batch_size=self.batch_size,
learning_rate=self.learning_rate,
B=1,
w_determinist=self.w_determinist,
init_params=init_params,
).learn(train_set)
self.Theta = np.array(pcrl_.Gs) / np.array(pcrl_.Gr)
self.Beta = np.array(pcrl_.Ls) / np.array(pcrl_.Lr)
# overwrite init_params for future fine-tuning
self.Gs = pcrl_.Gs
self.Gr = pcrl_.Gr
self.Ls = pcrl_.Ls
self.Lr = pcrl_.Lr
elif self.verbose:
print("%s is trained already (trainable = False)" % (self.name))
return self
[docs]
def score(self, user_idx, item_idx=None):
"""Predict the scores/ratings of a user for a list of items.
Parameters
----------
user_idx: int, required
The index of the user for whom to perform score prediction.
item_idx: int, optional, default: None
The index of the item for which to perform score prediction.
If None, scores for all known items will be returned.
Returns
-------
res : A scalar or a Numpy array
Relative scores that the user gives to the item or to all known items
"""
if item_idx is None:
user_pred = self.Beta * self.Theta[user_idx, :].T
else:
user_pred = self.Beta[item_idx, :] * self.Theta[user_idx, :].T
# transform user_pred to a flatten array
user_pred = np.array(user_pred, dtype="float64").flatten()
return user_pred
[docs]
def get_vector_measure(self):
"""Getting a valid choice of vector measurement in ANNMixin._measures.
Returns
-------
measure: MEASURE_DOT
Dot product aka. inner product
"""
return MEASURE_DOT
[docs]
def get_user_vectors(self):
"""Getting a matrix of user vectors serving as query for ANN search.
Returns
-------
out: numpy.array
Matrix of user vectors for all users available in the model.
"""
return self.Theta
[docs]
def get_item_vectors(self):
"""Getting a matrix of item vectors used for building the index for ANN search.
Returns
-------
out: numpy.array
Matrix of item vectors for all items available in the model.
"""
return self.Beta
