{"id":"W2899079481","doi":"10.1007/s11053-018-9423-2","title":"Surficial and Deep Earth Material Prediction from Geochemical Compositions","year":2018,"lang":"en","type":"article","venue":"Natural Resources Research","topic":"Geochemistry and Geologic Mapping","field":"Computer Science","cited_by":37,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Deutscher Akademischer Austauschdienst; Australian Government; Department for Enterprise, Trade and Investment, UK Government","keywords":"Regolith; Compositional data; Geology; Random forest; Footprint; Multivariate statistics; Mineral exploration; Scale (ratio); Earth science; Computer science; Artificial intelligence; Cartography; Geophysics; Machine learning","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":[0.0004676148,0.00009735698,0.0001117734,0.00008455776,0.0005180509,0.0002974232,0.0005565983,0.0001412351,0.0001794678],"category_scores_gemma":[0.0002831906,0.00008382551,0.00002810349,0.0003333651,0.0004228963,0.0001720217,0.000636253,0.0004681175,0.00005478884],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002211099,"about_ca_system_score_gemma":0.00002481396,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000189001,"about_ca_topic_score_gemma":0.00003998945,"domain_scores_codex":[0.9982792,0.0001766007,0.0001512248,0.0004308941,0.0005476139,0.0004144868],"domain_scores_gemma":[0.9989208,0.0002602655,0.00002758724,0.0003140082,0.0003505566,0.0001267945],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0003548303,0.0001168163,0.004419827,0.00003650812,0.00005413724,0.00005589717,0.003811025,0.00000947962,0.9410356,0.004667999,0.007037648,0.03840021],"study_design_scores_gemma":[0.001033193,0.0005283581,0.07632157,0.0001020567,0.00001022372,0.00009331312,0.0003611349,0.1265372,0.3966531,0.01950237,0.3783697,0.0004878176],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9847942,0.0001950404,0.0005271342,0.002040452,0.0002897555,0.000113368,0.00002408161,0.0001110657,0.01190487],"genre_scores_gemma":[0.9955834,0.00001067271,0.002475883,0.00005232507,0.001015614,0.000007788286,0.00004008313,0.000002044837,0.0008121575],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5443825,"threshold_uncertainty_score":0.3984481,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02635488771249922,"score_gpt":0.2987460791104011,"score_spread":0.2723911913979019,"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."}}