ALS ImplicitMF on ML100K

This page analyzes the hyperparameter tuning results for implicit-feedback matrix factorization with ALS.

Parameter Search Space

Parameter	Type	Distribution	Values
embedding_size	Integer	LogUniform	4 ≤ \(x\) ≤ 512
regularization.user	Float	LogUniform	1e-05 ≤ \(x\) ≤ 1
regularization.item	Float	LogUniform	1e-05 ≤ \(x\) ≤ 1
damping.user	Float	LogUniform	1e-12 ≤ \(x\) ≤ 100
damping.item	Float	LogUniform	1e-12 ≤ \(x\) ≤ 100

Final Result

Searching selected the following configuration:

{
    'embedding_size': 22,
    'regularization': {'user': 1.5567275542968303e-05, 'item': 0.00011227060161613435},
    'damping': {'user': 6.591894907169958e-08, 'item': 5.971049133063153e-08},
    'weight': 5.712005519451577,
    'epochs': 8
}

With these metrics:

{
    'RBP': 0.1344754176992636,
    'NDCG': 0.4091118573760802,
    'RecipRank': 0.3995648213557018,
    'TrainTask': '3910328d-b13e-4bea-8e9e-984e71d59f0b',
    'TrainTime': None,
    'TrainCPU': None,
    'timestamp': 1743029399,
    'checkpoint_dir_name': None,
    'done': True,
    'training_iteration': 8,
    'trial_id': '4e859_00089',
    'date': '2025-03-26_18-49-59',
    'time_this_iter_s': 0.18914508819580078,
    'time_total_s': 1.892554759979248,
    'pid': 269870,
    'hostname': 'CCI-ws21',
    'node_ip': '10.248.127.152',
    'config': {
        'embedding_size': 22,
        'regularization': {'user': 1.5567275542968303e-05, 'item': 0.00011227060161613435},
        'damping': {'user': 6.591894907169958e-08, 'item': 5.971049133063153e-08},
        'weight': 5.712005519451577,
        'epochs': 8
    },
    'time_since_restore': 1.892554759979248,
    'iterations_since_restore': 8,
    'experiment_tag': '89_item=0.0000,user=0.0000,embedding_size=22,item=0.0001,user=0.0000,weight=5.7120'
}

Iteration Completion

How many iterations, on average, did we complete?

How did the metric progress in the best result?

How did the metric progress in the longest results?