{"id":"W4415997694","doi":"10.2139/ssrn.5713643","title":"Multi-Task Learning for Customer Base Analysis: Evidence from 966 Companies","year":2025,"lang":"","type":"preprint","venue":"SSRN Electronic Journal","topic":"Customer churn and segmentation","field":"Business, Management and Accounting","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Quest University Canada","funders":"","keywords":"Customer base; Valuation (finance); Revenue; Benchmark (surveying); Cash flow; Order (exchange); Upstream (networking); Benchmarking; Downstream (manufacturing)","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":["metaepi_narrow","sts","scholarly_communication","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.005391344,0.001127826,0.001695611,0.002650801,0.001830835,0.00240222,0.001535847,0.0005321159,0.0006200937],"category_scores_gemma":[0.0008904082,0.001157583,0.002004414,0.002302563,0.0001404906,0.002248398,0.0008637346,0.007549409,0.0003488615],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002503777,"about_ca_system_score_gemma":0.003854906,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004365467,"about_ca_topic_score_gemma":0.0127999,"domain_scores_codex":[0.9910016,0.0002774713,0.001919444,0.001535125,0.001016192,0.004250142],"domain_scores_gemma":[0.9945132,0.0006749885,0.002696772,0.0006833214,0.001344329,0.00008737615],"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":[0.005291286,0.002427981,0.2697967,0.003618072,0.0797328,0.00005527817,0.005202717,0.4269324,0.005530819,0.01459588,0.002961666,0.1838544],"study_design_scores_gemma":[0.008586546,0.0002220844,0.02449941,0.004134592,0.05154558,0.00002334434,0.03621593,0.8187941,0.0001226775,0.02493839,0.02705097,0.003866407],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2264382,0.05318805,0.7089446,0.003257153,0.004391782,0.002610738,0.00008831579,0.0003024602,0.0007787295],"genre_scores_gemma":[0.9593321,0.02304832,0.001224366,0.0006009326,0.003117657,0.0001118388,0.0006891664,0.00009157739,0.01178402],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7328939,"threshold_uncertainty_score":0.9994686,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03205755231945347,"score_gpt":0.2913784864058122,"score_spread":0.2593209340863588,"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."}}