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 48
min_nbrs Integer Uniform 1 ≤ \(x\) ≤ 5 3
min_sim Float LogUniform 1e-06 ≤ \(x\) ≤ 0.1 0.000322

Final Result

Searching selected the following configuration:

{'max_nbrs': 48, 'min_nbrs': 3, 'min_sim': 0.00032219628627765763}

With these metrics:

{
    'RBP': 0.10934313659194403,
    'LogRBP': 1.5809756704367954,
    'NDCG': 0.3602513935293,
    'RecipRank': 0.36683737588977156,
    'TrainTask': '95ed7cf4-47e9-4351-95e0-ed5d43b525f9',
    'TrainTime': 0.034394936985336244,
    'TrainCPU': 0.025966999999999935,
    'TestTask': '5b9a41f4-c925-424d-99d7-8af99e024b0f',
    'TestTime': 18.842595135967713,
    'TestCPU': 18.888780999999998,
    'timestamp': 1746427543,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 1,
    'trial_id': '3b058e55',
    'date': '2025-05-05_02-45-43',
    'time_this_iter_s': 19.42402458190918,
    'time_total_s': 19.42402458190918,
    'pid': 2955752,
    'hostname': 'CCI-ws21',
    'node_ip': '10.248.127.152',
    'config': {'max_nbrs': 48, 'min_nbrs': 3, 'min_sim': 0.00032219628627765763},
    'time_since_restore': 19.42402458190918,
    'iterations_since_restore': 1,
    'experiment_tag': '29_max_nbrs=48,min_nbrs=3,min_sim=0.0003'
}

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.