UKNN Explicit

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

Parameter Search Space

Parameter Type Distribution Values Selected
max_nbrs Integer Uniform 2 ≤ \(x\) ≤ 50 48
min_nbrs Integer Uniform 1 ≤ \(x\) ≤ 5 4
min_sim Float LogUniform 1e-06 ≤ \(x\) ≤ 0.1 0.000156

Final Result

Searching selected the following configuration:

{'max_nbrs': 48, 'min_nbrs': 4, 'min_sim': 0.00015621684353078706}

With these metrics:

{
    'RBP': 0.010937867967888624,
    'NDCG': 0.17721284067191045,
    'RecipRank': 0.04424932763848633,
    'RMSE': 0.8523203742671881,
    'TrainTask': '849fed59-dc95-4140-9f20-454bc6e59ade',
    'TrainTime': 0.2349925369999255,
    'TrainCPU': 0.24175800000000028,
    'TestTask': '9c34e429-668d-4fad-a19b-291248da1735',
    'TestTime': 58.29620637500011,
    'TestCPU': 62.029911999999996,
    'timestamp': 1742697538,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 1,
    'trial_id': 'ac264_00022',
    'date': '2025-03-22_22-38-58',
    'time_this_iter_s': 59.310041189193726,
    'time_total_s': 59.310041189193726,
    'pid': 172881,
    'hostname': 'CCI-ws21',
    'node_ip': '10.248.127.152',
    'config': {'max_nbrs': 48, 'min_nbrs': 4, 'min_sim': 0.00015621684353078706},
    'time_since_restore': 59.310041189193726,
    'iterations_since_restore': 1,
    'experiment_tag': '22_max_nbrs=48,min_nbrs=4,min_sim=0.0002'
}

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.