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 13
min_nbrs Integer Uniform 1 ≤ \(x\) ≤ 5 4
min_sim Float LogUniform 1e-06 ≤ \(x\) ≤ 0.1 0.0487

Final Result

Searching selected the following configuration:

{'max_nbrs': 13, 'min_nbrs': 4, 'min_sim': 0.048732774977812005}

With these metrics:

{
    'RBP': 0.13904679136677653,
    'LogRBP': 1.8212951958116026,
    'NDCG': 0.41419387583214873,
    'RecipRank': 0.45199601094730585,
    'TrainTask': '8767b7b4-4d83-4838-9a0f-b5c36522f9a5',
    'TrainTime': 0.01884700800292194,
    'TrainCPU': 0.020484000000000502,
    'TestTask': '96b08e14-3404-4dac-831d-569d0d5e2902',
    'TestTime': 0.5086141760111786,
    'TestCPU': 0.5130479999999996,
    'timestamp': 1746583645,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 1,
    'trial_id': 'edee3a80',
    'date': '2025-05-06_22-07-25',
    'time_this_iter_s': 0.6637563705444336,
    'time_total_s': 0.6637563705444336,
    'pid': 766901,
    'hostname': 'gracehopper1',
    'node_ip': '192.168.225.60',
    'config': {'max_nbrs': 13, 'min_nbrs': 4, 'min_sim': 0.048732774977812005},
    'time_since_restore': 0.6637563705444336,
    'iterations_since_restore': 1,
    'experiment_tag': '59_max_nbrs=13,min_nbrs=4,min_sim=0.0487'
}

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.