{"id":"W4409124103","doi":"10.22541/au.174371573.35145749/v1","title":"Clustering analysis of very large measurement and model datasets on high performance computing platforms","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Simulation Techniques and Applications","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Environment and Climate Change Canada","funders":"","keywords":"Cluster analysis; Computer science; Data mining; Data science; Artificial intelligence","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003484191,0.001286048,0.0009864697,0.00330946,0.001398447,0.001655819,0.001427932,0.0007918548,0.00177384],"category_scores_gemma":[0.01624512,0.000535524,0.001521912,0.005591723,0.0005959353,0.001367161,0.001392893,0.001294672,0.0009303333],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001524084,"about_ca_system_score_gemma":0.002537992,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.03209997,"about_ca_topic_score_gemma":0.03003617,"domain_scores_codex":[0.9969184,0.0009180516,0.000260293,0.0006298302,0.001047069,0.0002262945],"domain_scores_gemma":[0.9915271,0.003107062,0.0005360796,0.002901769,0.001725145,0.0002029145],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004138532,0.0003953446,0.04712946,0.0006940175,0.00132254,0.0005818143,0.0007644115,0.7248843,0.01359369,0.01624793,0.03638248,0.15759],"study_design_scores_gemma":[0.00004875397,0.00006686687,0.03079236,0.00004443424,0.00005332566,0.00009977958,0.0004708088,0.9328988,0.006194049,0.01976186,0.009503067,0.00006588748],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2778783,0.0005573113,0.6764989,0.001136341,0.0003054672,0.0006276573,0.02291649,0.01512538,0.004954197],"genre_scores_gemma":[0.533652,0.0003985736,0.4239067,0.0001672132,0.00009295168,0.0008365771,0.03865326,0.000773503,0.001519197],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.03209997,"threshold_uncertainty_score":0.06382626,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.188099513850231,"score_gpt":0.416504869300075,"score_spread":0.228405355449844,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}