{"id":"W3163328251","doi":"10.2196/26540","title":"Predicting Depression From Smartphone Behavioral Markers Using Machine Learning Methods, Hyperparameter Optimization, and Feature Importance Analysis: Exploratory Study","year":2021,"lang":"en","type":"article","venue":"JMIR mhealth and uhealth","topic":"Digital Mental Health Interventions","field":"Psychology","cited_by":148,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Depression (economics); Hyperparameter; Patient Health Questionnaire; Machine learning; Psychology; mHealth; Clinical psychology; Mental health; Feature (linguistics); Artificial intelligence; Computer science; Applied psychology; Psychiatry; Psychological intervention; Cognition; Depressive symptoms","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003637115,0.0009077494,0.0005238606,0.0008325787,0.000237747,0.0005200352,0.0006635367,0.0005454632,0.0004936928],"category_scores_gemma":[0.01073609,0.0003245973,0.001205575,0.0005191366,0.0002861703,0.0005399295,0.0006576944,0.0006641254,0.000161366],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000370483,"about_ca_system_score_gemma":0.0005316274,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003934204,"about_ca_topic_score_gemma":0.003240071,"domain_scores_codex":[0.9987183,0.0008360202,0.00006744079,0.000206223,0.0001004315,0.00007164112],"domain_scores_gemma":[0.9933831,0.004986766,0.0004868763,0.0004737918,0.0005211466,0.0001482803],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001561978,0.002969007,0.7704172,0.0002142361,0.0007756077,0.0004112691,0.000534096,0.1029844,0.003648415,0.0006086047,0.001417865,0.1144574],"study_design_scores_gemma":[0.00008635793,0.001309051,0.1183286,0.00002750186,0.0001475276,0.0001608868,0.0002376578,0.8772593,0.00150977,0.0006226061,0.0002776604,0.00003311175],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9874238,0.0001260865,0.01191973,0.00006537392,0.000005625942,0.00007996339,0.0001946344,0.00004280397,0.0001418796],"genre_scores_gemma":[0.9860302,0.00006164266,0.01326473,0.00002066989,0.00001143603,0.00009819985,0.0003996199,0.000004755358,0.0001088559],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.003934204,"threshold_uncertainty_score":0.01923513,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09229724354807968,"score_gpt":0.4657507465112854,"score_spread":0.3734535029632057,"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."}}