UKNN Implicit

This page analyzes the hyperparameter tuning results for the ItemKNN scorer in implicit-feedback mode.

Parameter Search Space

Parameter Type Distribution Values Selected
max_nbrs Integer Uniform 2 ≤ \(x\) ≤ 50 17
min_nbrs Integer Uniform 1 ≤ \(x\) ≤ 5 4
min_sim Float LogUniform 1e-06 ≤ \(x\) ≤ 0.1 0.0198

Final Result

Searching selected the following configuration:

{'max_nbrs': 17, 'min_nbrs': 4, 'min_sim': 0.019837909637114078}

With these metrics:

{
    'RBP': 0.1388285900825859,
    'LogRBP': 1.8197246980733635,
    'NDCG': 0.4149488981748211,
    'RecipRank': 0.4398772016392619,
    'TrainTask': '492d1ad4-b3e6-490e-b9f5-5cf3bbe4c6aa',
    'TrainTime': 0.0181006719940342,
    'TrainCPU': 0.01960699999999954,
    'TestTask': 'abc1ecc3-0344-47fb-b318-fdc095d08c81',
    'TestTime': 0.5153395839734003,
    'TestCPU': 0.5213629999999998,
    'timestamp': 1746581089,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 1,
    'trial_id': '0cab5_00007',
    'date': '2025-05-06_21-24-49',
    'time_this_iter_s': 0.6660077571868896,
    'time_total_s': 0.6660077571868896,
    'pid': 549612,
    'hostname': 'gracehopper1',
    'node_ip': '192.168.225.60',
    'config': {'max_nbrs': 17, 'min_nbrs': 4, 'min_sim': 0.019837909637114078},
    'time_since_restore': 0.6660077571868896,
    'iterations_since_restore': 1,
    'experiment_tag': '7_max_nbrs=17,min_nbrs=4,min_sim=0.0198'
}

Parameter Behavior

Neighborhood Size

The neighborhood size is key argument to consider, let’s look at RMSE with repspect to that:

Filtering

There are two filtering parameters we also test — minimum neighbors and mininmum simialrity.