{"id":"W4384210660","doi":"10.1039/d2dd00149g","title":"Tackling data scarcity with transfer learning: a case study of thickness characterization from optical spectra of perovskite thin films","year":2023,"lang":"en","type":"article","venue":"Digital Discovery","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Science and Engineering Research Council; Energy Market Authority of Singapore; National Research Foundation Singapore; Agency for Science, Technology and Research; Natural Resources Canada; Massachusetts Institute of Technology; National Research Foundation; Canadian Natural Resources Limited; National University of Singapore","keywords":"Reflection (computer programming); Transfer of learning; Characterization (materials science); Perovskite (structure); Spectral line; Data transmission; Materials science; Scarcity; Thin film; Transfer (computing); Transmission (telecommunications); Computer science; Optics; Chemical engineering; Nanotechnology; Artificial intelligence; Physics; Engineering; Telecommunications; Economics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.000687999,0.0002098248,0.0004166721,0.0001081001,0.0001233774,0.000552105,0.0007117926,0.00005832366,0.0001192991],"category_scores_gemma":[0.0003548313,0.0001630739,0.00003811836,0.0004832629,0.0002538035,0.00270683,0.0004153497,0.0001775752,0.00003596505],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001295232,"about_ca_system_score_gemma":0.00008869018,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000555566,"about_ca_topic_score_gemma":0.00006092662,"domain_scores_codex":[0.9976867,0.0001511738,0.000505844,0.000671387,0.0006846942,0.0003001791],"domain_scores_gemma":[0.9986595,0.000266513,0.0001688638,0.0007535683,0.00007995629,0.00007160801],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.0002226655,0.0004277507,0.03973266,0.00009706686,0.00002295248,0.0004723362,0.005182629,0.01417386,0.9393275,0.000116922,0.000005093643,0.0002185834],"study_design_scores_gemma":[0.005034158,0.004191062,0.4390693,0.0009551721,0.000358931,0.0007140712,0.04517636,0.1696117,0.3323578,0.0003291102,0.00006782269,0.002134487],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9904346,0.000003589287,0.007739275,0.00002573832,0.0002058521,0.0003315656,0.0009494073,0.0001256112,0.0001843267],"genre_scores_gemma":[0.998826,0.00000218759,0.0003736558,0.000008583961,0.0000505231,0.000008993222,0.0005840565,0.00002956594,0.000116481],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6069697,"threshold_uncertainty_score":0.664996,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02620805371855942,"score_gpt":0.2703795894013383,"score_spread":0.2441715356827789,"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."}}