Innovation science and technologiy 2-том 5-нөмір (2026)
METHODOLOGICAL FOUNDATIONS FOR FORMING CUSTOMS RISK PROFILE INDICATORS BASED ON MATHEMATICAL AND STATISTICAL METHODS
Niyazov, Sherzod
Аңдатпа
The paper investigates a mathematical and statistical methodology for designing risk profileindicators (RPIs) — the central element of customs risk management. Using two datasets — 16,035consignments with detected discrepancies and 616,925 declarations — seven base indicators are combinedinto 120 combinations and 386,516 risk indicator sets. Conditional probabilities are computed with Bayes–Laplace smoothing, and the lift coefficient ranks combinations by their fraud-detection power.
risk management, customs control, risk profile, conditional probability, Laplace smoothing, lift coefficient, naive Bayes classifier, association rules
Метадеректер дереккөзі: журналдың OAI-PMH архиві · Sindex толық мәтінді сақтамайды, дереккөзге сілтеме береді.