{"id":"W4416242726","doi":"10.1016/j.cities.2025.106429","title":"Contextualized poverty targeting with multimodal spatial data and machine learning in Brazzaville, Congo","year":2025,"lang":"en","type":"article","venue":"Cities","topic":"Impact of Light on Environment and Health","field":"Environmental Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Microsoft Azure; Faculté de Biologie et de Médecine, Université de Lausanne; Microsoft","keywords":"Poverty; Leverage (statistics); Human welfare; Capability approach; Welfare; Data collection; Urbanization; Field (mathematics)","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0003227095,0.0001187743,0.0001725412,0.00003881739,0.0001531722,0.00002937986,0.0001729098,0.00003056502,0.0009672331],"category_scores_gemma":[0.00006746546,0.00009399048,0.000009163184,0.00007242927,0.0002136711,0.0002481621,0.0003261163,0.000177893,0.00001942502],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007850077,"about_ca_system_score_gemma":0.00001585022,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004991965,"about_ca_topic_score_gemma":0.001974677,"domain_scores_codex":[0.9990717,0.00006655547,0.0001672659,0.0002635745,0.0001514971,0.0002794613],"domain_scores_gemma":[0.9996219,0.00008506725,0.00005269099,0.0001781772,0.000001482095,0.00006063887],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001011832,0.0000278025,0.9931201,0.00001667454,0.0000072807,0.000008147274,0.0006486078,0.00006666366,0.0005254039,0.00007078022,0.0006430552,0.004764338],"study_design_scores_gemma":[0.002374784,0.0001344162,0.9335712,0.00005881101,0.00001270998,0.000004169096,0.001214632,0.01500798,0.0003074225,0.00009788992,0.0469829,0.0002331415],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9821951,0.0003392226,0.0006846545,0.0009954124,0.00004811499,0.0001868711,0.00002476255,0.00003619423,0.01548964],"genre_scores_gemma":[0.9958661,0.0001219169,0.0005244004,0.0008368393,0.00001530843,0.000004442795,0.00007164261,0.000008802579,0.002550592],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.05954892,"threshold_uncertainty_score":0.999946,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01664809852084799,"score_gpt":0.2536286943273071,"score_spread":0.2369805958064591,"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."}}