{"id":"W4411409968","doi":"10.1371/journal.pone.0325321","title":"Incorporating sparse labels into hidden Markov models using weighted likelihoods improves accuracy and interpretability in biologging studies","year":2025,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Marine animal studies overview","field":"Environmental Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University; University of British Columbia","funders":"Canada Excellence Research Chairs, Government of Canada; University of British Columbia Graduate School; British Columbia Knowledge Development Fund; Natural Sciences and Engineering Research Council of Canada; Fisheries and Oceans Canada; Canada Foundation for Innovation","keywords":"Hidden Markov model; Computer science; Interpretability; Artificial intelligence; Process (computing); Machine learning; Markov chain; Markov model; Mark and recapture; Frame (networking); Pattern recognition (psychology)","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01319921,0.00153527,0.001265679,0.002127424,0.0008485682,0.002891435,0.002026424,0.002341038,0.002034555],"category_scores_gemma":[0.07277316,0.0007776197,0.001451223,0.001239953,0.001089802,0.004982433,0.002614402,0.00373618,0.0008682212],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001313836,"about_ca_system_score_gemma":0.001843747,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01098556,"about_ca_topic_score_gemma":0.01724989,"domain_scores_codex":[0.9952286,0.002935974,0.0003047828,0.0009073613,0.0004871865,0.000136158],"domain_scores_gemma":[0.9367887,0.05488816,0.002648459,0.003047844,0.002163333,0.0004635255],"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":[0.0006347599,0.0002548071,0.04407749,0.0005279173,0.0004752532,0.0005239092,0.001360936,0.6384929,0.005320693,0.02241995,0.004086101,0.2818254],"study_design_scores_gemma":[0.00002112553,0.00003940928,0.001384277,0.00005828304,0.0000355229,0.00003847167,0.00007086097,0.9703902,0.0009996822,0.02614832,0.0007874584,0.0000264238],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03806375,0.0004801846,0.9577938,0.0007725726,0.00006394517,0.00006537842,0.0004002542,0.001286861,0.001073204],"genre_scores_gemma":[0.6016167,0.0005853659,0.392877,0.0004885705,0.0001736479,0.0002426051,0.001848539,0.0004608936,0.001706576],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01319921,"threshold_uncertainty_score":0.06980491,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06080760260929913,"score_gpt":0.2883654003852599,"score_spread":0.2275577977759607,"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."}}