{"id":"W4220747158","doi":"10.1136/bmjsem-2021-001259","title":"Why machine learning (ML) has failed physical activity research and how we can improve","year":2022,"lang":"en","type":"article","venue":"BMJ Open Sport & Exercise Medicine","topic":"Physical Activity and Health","field":"Medicine","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan; University of Calgary; Memorial University of Newfoundland","funders":"","keywords":"Benchmark (surveying); Computer science; Field (mathematics); Software; Physical activity; Point (geometry); Measure (data warehouse); Machine learning; Artificial intelligence; Data science; Human–computer interaction; Data mining; Physical medicine and rehabilitation; Medicine","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.1545361,0.001781293,0.00354026,0.006075264,0.002398228,0.01338675,0.00506828,0.00684035,0.006977219],"category_scores_gemma":[0.3935994,0.001273634,0.002745693,0.005566302,0.01619873,0.02619784,0.005330147,0.01288304,0.005977238],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005580399,"about_ca_system_score_gemma":0.01003544,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009450875,"about_ca_topic_score_gemma":0.006740815,"domain_scores_codex":[0.9160736,0.05092017,0.006286772,0.008405329,0.01653209,0.001781929],"domain_scores_gemma":[0.4842789,0.4267713,0.01316642,0.02318832,0.04738255,0.005212485],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0003141642,0.0002643906,0.01939149,0.008391351,0.000923204,0.0001383545,0.002567799,0.003305431,0.0006017575,0.11028,0.1255741,0.7282479],"study_design_scores_gemma":[0.0002190623,0.0006343035,0.01275992,0.01800867,0.0007055837,0.0004758083,0.002778788,0.0108194,0.002264097,0.6333023,0.3175776,0.0004544445],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"commentary","genre_gemma":"empirical","genre_scores_codex":[0.007241265,0.1370439,0.0987175,0.7310121,0.009587927,0.0001853141,0.0009993354,0.001632567,0.01358013],"genre_scores_gemma":[0.2381206,0.1587618,0.2765793,0.2856188,0.02496762,0.001494797,0.001929512,0.003330583,0.009197057],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8454639,"threshold_uncertainty_score":0.817275,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1753162058630499,"score_gpt":0.4336553102622821,"score_spread":0.2583391043992322,"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."}}