{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002460934,0.0001661798,0.0004901033,0.0008019643,0.0001427665,0.0001121693,0.0005993327,0.0001266588,0.00005859646],"category_scores_gemma":[0.0001525874,0.0001220174,0.0001196971,0.000774115,0.00002845456,0.00008882843,0.001419124,0.0002017571,0.000003312335],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006171148,"about_ca_system_score_gemma":0.00006955562,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006107485,"about_ca_topic_score_gemma":0.00005563307,"domain_scores_codex":[0.9972939,0.00001765358,0.0007914714,0.0005821288,0.001167717,0.0001471466],"domain_scores_gemma":[0.997936,0.0002722764,0.0003760411,0.001058434,0.0003088575,0.0000483988],"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":[0.00001215444,0.00004593624,0.002351116,0.00002756767,0.0001417318,7.280973e-8,0.00008049628,0.9770972,0.00001618184,0.003060592,0.0008609186,0.01630603],"study_design_scores_gemma":[0.0001063605,0.00001056266,0.01077829,0.00007124102,0.0001471964,5.44368e-8,0.00003773305,0.9844847,0.0004750786,0.003576104,0.0001933363,0.0001193263],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3119139,0.00001901124,0.6850888,0.00006638321,0.00004373966,0.0002461332,0.0008508914,0.0000520794,0.001719041],"genre_scores_gemma":[0.9856476,0.00003243297,0.01370276,0.0001650994,0.00001078911,0.00001332126,0.0002444579,0.00000398522,0.0001795174],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6737338,"threshold_uncertainty_score":0.4975725,"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."}}