Point process modeling of drug overdoses with heterogeneous and missing data
نویسندگان
چکیده
Opioid overdose rates have increased in the United States over past decade and reflect a major public health crisis. Modeling prediction of drug opioid hotspots, where high percentage events fall small space–time, could help better focus limited social services. In this work we present spatial-temporal point process model for clustering. The data input into comes from two heterogeneous sources: (1) volume emergency medical calls service (EMS) records containing location time but no information on type nonfatal overdose, (2) fatal toxicology reports coroner high-dimensional screen drugs at death. We first use nonnegative matrix factorization to cluster categories, then develop an EM algorithm integrating sets, mark corresponding category is inferred EMS used more accurately predict death hotspots. apply Indianapolis, showing that defined integrated out-performs processes only (AUC improvement 0.81 0.85). also investigate extent which overdoses are contagious, as function while controlling exogenous fluctuations background rate might contribute find deaths exhibit significant excitation with branching ratio ranging 0.72 0.98.
منابع مشابه
Facies Modeling of Heterogeneous Carbonates Reservoirs by Multiple Point Geostatistics
Facies modeling is an essential part of reservoir characterization. The connectivity of facies model is very critical for the dynamic modeling of reservoirs. Carbonate reservoirs are so heterogeneous that variogram-based methods like sequential indicator simulation are not very useful for facies modeling. In this paper, multiple point geostatistics (MPS) is used for facies modeling in one of th...
متن کاملFailure Process Modeling with Censored Data in Accelerated Life Tests
Manufacturers need to evaluate the reliability of their products in order to increase the customer satisfaction. Proper analysis of reliability also requires an effective study of the failure process of a product, especially its failure time. So, the Failure Process Modeling (FPM) plays a key role in the reliability analysis of the system that has been less focused on. This paper introduces a f...
متن کاملDEA with Missing Data: An Interval Data Assignment Approach
In the classical data envelopment analysis (DEA) models, inputs and outputs are assumed as known variables, and these models cannot deal with unknown amounts of variables directly. In recent years, there are few researches on handling missing data. This paper suggests a new interval based approach to apply missing data, which is the modified version of Kousmanen (2009) approach. First, the prop...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: The Annals of Applied Statistics
سال: 2021
ISSN: ['1941-7330', '1932-6157']
DOI: https://doi.org/10.1214/20-aoas1384