IKNN 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 15
min_nbrs Integer Uniform 1 ≤ \(x\) ≤ 5 3
min_sim Float LogUniform 1e-06 ≤ \(x\) ≤ 0.1 0.0127

Final Result

Searching selected the following configuration:

{'max_nbrs': 15, 'min_nbrs': 3, 'min_sim': 0.012663408844327352}

With these metrics:

{
    'RBP': 0.009878131828519423,
    'LogRBP': -0.8231919015118887,
    'NDCG': 0.18895378376288843,
    'RecipRank': 0.0453315358348199,
    'RMSE': 0.9170927975385909,
    'TrainTask': '85592e1b-0298-4dd6-bca5-a78b294a9028',
    'TrainTime': 0.05831209599273279,
    'TrainCPU': 0.0697810000000002,
    'TestTask': 'e436cce1-a5ce-449f-b695-ca5d38b1da4e',
    'TestTime': 2.1647912000189535,
    'TestCPU': 7.395099,
    'timestamp': 1746654765,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 1,
    'trial_id': '81ef5_00038',
    'date': '2025-05-07_17-52-45',
    'time_this_iter_s': 2.38818359375,
    'time_total_s': 2.38818359375,
    'pid': 1420270,
    'hostname': 'gracehopper1',
    'node_ip': '192.168.225.60',
    'config': {'max_nbrs': 15, 'min_nbrs': 3, 'min_sim': 0.012663408844327352},
    'time_since_restore': 2.38818359375,
    'iterations_since_restore': 1,
    'experiment_tag': '38_max_nbrs=15,min_nbrs=3,min_sim=0.0127'
}

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.