def model(train_interactions_ds):
item_knn = UserKNN(k=3,
m=0,
sim_metric='cosine',
aggregation='weighted_mean',
shrinkage=100,
use_averages=False)
item_knn.fit(train_interactions_ds, verbose=False)
return item_knn
ds = InteractionDataset('./cheRM_total.csv',
columns=['user', 'item', 'interaction'],
verbose=False)
ds_train, ds_test = matrix_split(ds,
min_user_interactions=20,
user_test_ratio=0.2,
item_test_ratio=0.2,
seed=25,
verbose=False)
# cosine sim
knn = UserKNN(k=10,
m=0,
sim_metric='cosine_cf',
shrinkage=None,
seed=25,
use_averages=False,
verbose=True)
knn.fit(ds_train)
evaluation = ranking_evaluation(knn,
ds_test,
interaction_threshold=2,
k=list(range(1, 11)),
generate_negative_pairs=False,
n_pos_interactions=None,
n_neg_interactions=None,
seed=25,
verbose=True,
metrics=[Precision(),
def fit_model_mean_aggr(train_interaction_ds):
fit_model_mean_aggr = UserKNN(k=20, m=5, sim_metric='adjusted_cosine', aggregation='mean', shrinkage=100,
use_averages=False)
fit_model_mean_aggr.fit(train_interaction_ds)
return fit_model_mean_aggr
def fit_model_cosine_sim(train_interaction_ds):
fit_model_cosine_sim = UserKNN(k=20, m=5, sim_metric='cosine', aggregation='weighted_mean', shrinkage=100,
use_averages=False)
fit_model_cosine_sim.fit(train_interaction_ds)
return fit_model_cosine_sim
def fit_model_no_shrinkage(train_interaction_ds):
fit_model_no_shrinkage = UserKNN(k=20, m=5, sim_metric='adjusted_cosine', aggregation='weighted_mean',
shrinkage=None, use_averages=False)
fit_model_no_shrinkage.fit(train_interaction_ds)
return fit_model_no_shrinkage
def fit_model_use_averages(train_interaction_ds):
fit_model_use_averages = UserKNN(k=1, m=1, sim_metric='adjusted_cosine', aggregation='weighted_mean', shrinkage=100,
use_averages=True)
fit_model_use_averages.fit(train_interaction_ds)
return fit_model_use_averages