{"id":"W4392255248","doi":"10.1145/3638209.3638224","title":"Occupancy Estimation in Smart Buildings: Impact of Data Quality on Feature Selection","year":2023,"lang":"en","type":"article","venue":"","topic":"Mobile Crowdsensing and Crowdsourcing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Computer science; Feature selection; Reliability (semiconductor); Machine learning; Data mining; Feature (linguistics); Selection (genetic algorithm); Occupancy; Generalization; Artificial intelligence; Quality (philosophy); Process (computing); Power (physics); Engineering","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009077344,0.00009455979,0.0001459272,0.0002190979,0.00005549756,0.00007725636,0.0004694682,0.00006326869,0.000005653038],"category_scores_gemma":[0.0001795061,0.00007823149,0.00004158451,0.00127355,0.00001381222,0.0005034533,0.0001803874,0.0001363621,0.00002661813],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006692149,"about_ca_system_score_gemma":0.00007086992,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008301514,"about_ca_topic_score_gemma":0.0001059338,"domain_scores_codex":[0.9989504,0.00008419131,0.0001975761,0.0003637656,0.0002100724,0.0001940188],"domain_scores_gemma":[0.9989806,0.0001405853,0.00008915966,0.0007074804,0.00004426421,0.00003792141],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001284584,0.0004006391,0.1761305,0.0002006763,0.00009859166,0.00002879668,0.0030504,0.08445653,0.04016057,0.02975853,0.06219107,0.6033953],"study_design_scores_gemma":[0.0002060269,0.00009138694,0.2488115,0.00006556344,0.00000213939,0.000007344298,0.00002522496,0.7469631,0.003053614,0.0005397947,0.0001168068,0.000117467],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8239783,0.000007662717,0.1742627,0.0003294443,0.0001027281,0.000102384,0.000005664148,0.0002836558,0.0009274582],"genre_scores_gemma":[0.9786695,0.000002664155,0.02100036,0.0000347036,0.00001692988,0.000001811593,0.00002627905,0.000005231261,0.0002425156],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6625066,"threshold_uncertainty_score":0.3190188,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06198769871351883,"score_gpt":0.3803489362575788,"score_spread":0.3183612375440599,"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."}}