MADSCAG: A Missingness-aware Early Warning System for Village-level Rice Pest Surveillance Under Reactive Monitoring
Year: 2026
Type: article
Abstract: Rice pest patrols go where farmers report trouble, so the register measures the observer as much as the pest.A forecast fitted to it inherits that bias.Is the observation process better modelled than corrected away?MADSCAG, missingness-aware decoupled stages for capacity allocation and guidance, predicts attendance and conditional severity separately on an East Java register of 114,757 village-periods, scored once chronologically.After a gap, officers found an attack 87.1% of the time against 43.1%; attendance reaches AUC 0.789 against 0.662 for a visit-rate rule; intervals cover 90.0 to 90.5% marginally, 2.1% on the largest outbreaks; the better-ranked half of visited villages holds 64.9% of recorded severity, 30% above random.No accuracy advantage over a recurrent alternative is demonstrated.Reactive registers make the visit predictable and informative, but no decision product built on it beat simpler rules, the blindspot list included; validating missed outbreaks needs prospective sampling.
Authors: Mike Yuliana(Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Supervision, Validation, Visualization, Writing – review & editing), Arthur Susilo(Conceptualization, Data curation, Formal analysis, Methodology, Software, Validation, Visualization, Writing – original draft), Riyadh Hadinah Ahtar Lakadimu(Conceptualization, Formal analysis, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing), Weny Mistarika Rahmawati(Funding acquisition, Project administration, Resources), Hendri Darmawan(Investigation, Validation, Writing – review & editing), Mochammad Zen Samsono Hadi(Resources, Supervision, Writing – review & editing), Rudy Dikairono(Project administration, Resources, Supervision), Gallyndra Fatkhu Dinata(Data curation, Investigation, Resources, Supervision, Validation), Rahardhita Widyatra Sudibyo(Data curation, Investigation, Resources)
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: hybrid
OpenAlex ID: W7235558317
