IKNN 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 49
min_nbrs Integer Uniform 1 ≤ \(x\) ≤ 5 3
min_sim Float LogUniform 1e-06 ≤ \(x\) ≤ 0.1 0.0298

Final Result

Searching selected the following configuration:

{'max_nbrs': 49, 'min_nbrs': 3, 'min_sim': 0.029797501413259136}

With these metrics:

{
    'RBP': 0.10970675789228801,
    'LogRBP': 1.5842956595670898,
    'NDCG': 0.3651186024846035,
    'RecipRank': 0.3874499674567859,
    'TrainTask': '44b10184-93ae-4185-809d-948fd62f50a2',
    'TrainTime': 0.045779040025081486,
    'TrainCPU': 0.047237000000000307,
    'TestTask': '215e5390-2ccf-42ff-bd5f-5ec1503844b2',
    'TestTime': 1.8532645440427586,
    'TestCPU': 8.186563,
    'timestamp': 1746655123,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 1,
    'trial_id': '45de3_00079',
    'date': '2025-05-07_17-58-43',
    'time_this_iter_s': 2.0157086849212646,
    'time_total_s': 2.0157086849212646,
    'pid': 1520142,
    'hostname': 'gracehopper1',
    'node_ip': '192.168.225.60',
    'config': {'max_nbrs': 49, 'min_nbrs': 3, 'min_sim': 0.029797501413259136},
    'time_since_restore': 2.0157086849212646,
    'iterations_since_restore': 1,
    'experiment_tag': '79_max_nbrs=49,min_nbrs=3,min_sim=0.0298'
}

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.