{"id":"W4414682428","doi":"10.26554/sti.2025.10.4.999-1011","title":"Decision Tree Algorithms in Water Quality Classification: A Comparative Study of Random Forest, XGBoost, and C5.0","year":2025,"lang":"en","type":"article","venue":"Science & Technology Indonesia","topic":"Water Quality Monitoring Technologies","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Decision tree; Random forest; Water quality; Preprocessor; Quality (philosophy); Random tree; Pipeline (software); Forage","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.001815115,0.0001873247,0.0004158072,0.00103576,0.0002677377,0.00003759734,0.001204222,0.0002012064,0.000007660971],"category_scores_gemma":[0.0001751593,0.0001370028,0.00002447436,0.00292871,0.003998608,0.000344688,0.001192959,0.0003016639,0.00001773493],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000224966,"about_ca_system_score_gemma":0.00003158343,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003392145,"about_ca_topic_score_gemma":0.0006377678,"domain_scores_codex":[0.9976437,0.0001007622,0.0005973866,0.0007454531,0.0004803906,0.0004322881],"domain_scores_gemma":[0.9988263,0.0001439041,0.0001302692,0.0008210127,0.00004208379,0.00003639383],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00007965151,0.0003358319,0.9408818,0.000004986344,0.000005847986,0.000004517895,0.00140307,0.00009824202,0.03687891,0.001264943,0.00001968514,0.01902248],"study_design_scores_gemma":[0.001917726,0.0002205023,0.9039938,0.0000301849,0.000008899326,0.000003842461,0.006588355,0.0007164423,0.06768025,0.01863563,0.00004683034,0.0001575615],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9961764,0.00003243414,0.001153995,0.001142044,0.0001319647,0.0007421605,9.807695e-7,0.0001936584,0.0004264289],"genre_scores_gemma":[0.9967172,0.000006491692,0.00308934,0.00001232975,0.000003385523,0.0001256379,5.546416e-7,0.000004200703,0.0000408994],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03688806,"threshold_uncertainty_score":0.9987119,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04732454753566626,"score_gpt":0.348229684982627,"score_spread":0.3009051374469607,"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."}}