def __init__(self, sess, config, dataset, evaluator):
super(CFGAN, self).__init__(sess, config, dataset, evaluator)
train_matrix = dataset.train_matrix
self.users_num, self.items_num = train_matrix.shape
self.g_lr = config["g_lr"]
self.d_lr = config["d_lr"]
self.d_hidden_units = config["d_hidden_units"]
self.g_hidden_units = config["g_hidden_units"]
self.epochs = config["epochs"]
self.batch_size = config["batch_size"]
self.s_zr = config["s_zr"]
self.s_pm = config["s_pm"]
self.alpha = config["alpha"]
self.user_pos_train = csr_to_user_dict(train_matrix)
self.all_items = np.arange(self.items_num)
self.evaluator = evaluator
self.train_matrix_dense = train_matrix.todense()
self.g_layers = None
self.d_layers = None
self._build_model()
self.sess = sess
self.sess.run(tf.global_variables_initializer())
def __init__(self):
super(APL, self).__init__()
train_matrix = self.dataset.train_matrix
self.users_num, self.items_num = train_matrix.shape
self.factors_num = self.config["factors_num"]
self.lr = self.config["lr"]
self.g_reg = self.config["g_reg"]
self.d_reg = self.config["d_reg"]
self.epochs = self.config["epochs"]
self.batch_size = self.config["batch_size"]
self.pretrain_file = self.config["pretrain_file"]
self.user_pos_train = csr_to_user_dict(train_matrix)
self.all_items = np.arange(self.items_num)
self._build_model()
self.sess.run(tf.global_variables_initializer())
def __init__(self):
super(CML, self).__init__()
train_matrix = self.dataset.train_matrix
self.users_num, self.items_num = train_matrix.shape
self.factors_num = self.config["factors_num"]
self.lr = self.config["lr"]
self.reg = self.config["reg"]
self.epochs = self.config["epochs"]
self.batch_size = self.config["batch_size"]
self.dns = self.config["dns"]
self.margin = self.config["margin"]
self.clip_norm = self.config["clip_norm"]
self.user_pos_train = csr_to_user_dict(train_matrix)
self.all_items = np.arange(self.items_num)
self.train_matrix = train_matrix
self._build_model()
self.sess.run(tf.global_variables_initializer())
def __init__(self, sess, config, dataset, evaluator):
super(APR, self).__init__(sess, config, dataset, evaluator)
train_matrix = dataset.train_matrix
self.users_num, self.items_num = train_matrix.shape
self.factors_num = config["factors_num"]
self.lr = config["lr"]
self.reg = config["reg"]
self.reg_adv = config["reg_adv"]
self.epochs = config["epochs"]
self.batch_size = config["batch_size"]
self.eps = config["eps"]
self.pretrain_file = config["pretrain_file"]
self.user_pos_train = csr_to_user_dict(train_matrix)
self.all_items = np.arange(self.items_num)
self.evaluator = evaluator
self._build_model()
self.sess = sess
self.sess.run(tf.global_variables_initializer())
def __init__(self):
super(APR, self).__init__()
train_matrix = self.dataset.train_matrix
self.num_users, self.num_items = train_matrix.shape
self.embedding_size = self.config["embed_size"]
self.learning_rate = self.config["lr"]
self.batch_size = self.config["batch_size"]
self.reg = self.config["reg"]
self.dns = self.config["dns"]
self.adv = self.config["adv"]
self.eps = self.config["eps"]
self.adv_epoch = self.config["adv_epoch"]
self.reg_adv = self.config["reg_adv"]
self.epochs = self.config["epochs"]
self.train_matrix = train_matrix
self.user_pos_train = csr_to_user_dict(train_matrix)
self.all_items = np.arange(self.num_items)
self.build_model()
self.sess.run(tf.global_variables_initializer())
def __init__(self):
super(CFGAN, self).__init__()
self.epochs = self.config["epochs"]
self.mode = self.config["mode"]
self.reg_G = self.config["reg_g"]
self.reg_D = self.config["reg_d"]
self.lr_G = self.config["lr_g"]
self.lr_D = self.config["lr_d"]
self.batchSize_G = self.config["batchsize_g"]
self.batchSize_D = self.config["batchsize_d"]
self.opt_G = self.config["opt_g"]
self.opt_D = self.config["opt_d"]
self.hiddenLayer_G = self.config["hiddenlayer_g"]
self.hiddenLayer_D = self.config["hiddenlayer_d"]
self.step_G = self.config["step_g"]
self.step_D = self.config["step_d"]
self.ZR_ratio = self.config["zr_ratio"]
self.ZP_ratio = self.config["zp_ratio"]
self.ZR_coefficient = self.config["zr_coefficient"]
self.train_matrix = self.dataset.train_matrix.copy()
# convert explicit data to implicit data
self.train_matrix.data[:] = 1
if self.mode == "itemBased":
self.train_matrix = self.train_matrix.transpose(copy=True).tocsr()
self.users_num, self.items_num = self.train_matrix.shape
self.user_pos_train = csr_to_user_dict(self.train_matrix)
self.all_items = np.arange(self.items_num)
self._build_model()
self.sess.run(tf.global_variables_initializer())