Nedialko B. Dimitrov

Optimizing tactics for use of the U.S. antiviral strategic national stockpile for pandemic influenza.

In 2009, public health agencies across the globe worked to mitigate the impact of the swine-origin influenza A (pH1N1) virus. These efforts included intensified surveillance, social distancing, hygiene measures, and the targeted use of antiviral medications to prevent infection (prophylaxis). In addition, aggressive antiviral treatment was recommended for certain patient subgroups to reduce the severity and duration of symptoms. To assist States and other localities meet these needs, the U.S. Government distributed a quarter of the antiviral medications in the Strategic National Stockpile within weeks of the pandemics start. However, there are no quantitative models guiding the geo-temporal distribution of the remainder of the Stockpile in relation to pandemic spread or severity. We present a tactical optimization model for distributing this stockpile for treatment of infected cases during the early stages of a pandemic like 2009 pH1N1, prior to the wide availability of a strain-specific vaccine. Our optimization method efficiently searches large sets of intervention strategies applied to a stochastic network model of pandemic influenza transmission within and among U.S. cities. The resulting optimized strategies depend on the transmissability of the virus and postulated rates of antiviral uptake and wastage (through misallocation or loss). Our results suggest that an aggressive community-based antiviral treatment strategy involving early, widespread, pro-rata distribution of antivirals to States can contribute to slowing the transmission of mildly transmissible strains, like pH1N1. For more highly transmissible strains, outcomes of antiviral use are more heavily impacted by choice of distribution intervals, quantities per shipment, and timing of shipments in relation to pandemic spread. This study supports previous modeling results suggesting that appropriate antiviral treatment may be an effective mitigation strategy during the early stages of future influenza pandemics, increasing the need for systematic efforts to optimize distribution strategies and provide tactical guidance for public health policy-makers.

Competitive Weighted Matching in Transversal Matroids

Consider a bipartite graph with a set of left-vertices and a set of right-vertices. All the edges adjacent to the same left-vertex have the same weight. We present an algorithm that, given the set of right-vertices and the number of left-vertices, processes a uniformly random permutation of the left-vertices, one left-vertex at a time. In processing a particular left-vertex, the algorithm either permanently matches the left-vertex to a thus-far unmatched right-vertex, or decides never to match the left-vertex. The weight of the matching returned by our algorithm is within a constant factor of that of a maximum weight matching.

Optimizing Provider Recruitment for Influenza Surveillance Networks

The increasingly complex and rapid transmission dynamics of many infectious diseases necessitates the use of new, more advanced methods for surveillance, early detection, and decision-making. Here, we demonstrate that a new method for optimizing surveillance networks can improve the quality of epidemiological information produced by typical provider-based networks. Using past surveillance and Internet search data, it determines the precise locations where providers should be enrolled. When applied to redesigning the provider-based, influenza-like-illness surveillance network (ILINet) for the state of Texas, the method identifies networks that are expected to significantly outperform the existing network with far fewer providers. This optimized network avoids informational redundancies and is thereby more effective than networks designed by conventional methods and a recently published algorithm based on maximizing population coverage. We show further that Google Flu Trends data, when incorporated into a network as a virtual provider, can enhance but not replace traditional surveillance methods.

Network deployment of radiation detectors with physics-based detection probability calculations

We describe a model for deploying radiation detectors on a transportation network consisting of two adversaries: a nuclear-material smuggler and an interdictor. The interdictor first installs the detectors. These installations are transparent to the smuggler, and are made under an uncertain threat scenario, which specifies the smuggler’s origin and destination, the nature of the material being smuggled, the manner in which it is shielded, and the mechanism by which the smuggler selects a route. The interdictor’s goal is to minimize the probability the smuggler evades detection. The performance of the detection equipment depends on the material being sensed, geometric attenuation, shielding, cargo and container type, background, time allotted for sensing and a number of other factors. Using a stochastic radiation transport code (MCNPX), we estimate detection probabilities for a specific set of such parameters, and inform the interdiction model with these estimates.

Optimal cover time for a graph-based coupon collector process

In this paper we study the following covering process defined over an arbitrary directed graph. Each node is initially uncovered and is assigned a random integer rank drawn from a suitable range. The process then proceeds in rounds. In each round, a uniformly random node is selected and its lowest-ranked uncovered outgoing neighbor, if any, is covered. We prove that if each node has in-degree

Combinatorial design of a stochastic Markov decision process

\theta({\it d})

Assessing real-time Zika risk in the United States

and out-degree O(d), then with high probability, every node is covered within

Interdiction Models and Applications

O(n+ \frac{n \ {\rm log}\ n}{d})

Convex Approximations of a Probabilistic Bicriteria Model with Disruptions

rounds, matching a lower bound due to Alon. Alon has also shown that, for a certain class of d-regular expander graphs, the upper bound holds no matter what method is used to choose the uncovered neighbor. In contrast, we show that for arbitrary d-regular graphs, the method used to choose the uncovered neighbor can affect the cover time by more than a constant factor.

Optimal feedback allocation algorithms for multi-user uplink

We consider a problem in which we seek to optimally design a Markov decision process (MDP). That is, subject to resource constraints we first design the action sets that will be available in each state when we later optimally control the process. The control policy is subject to additional constraints governing state-action pair frequencies, and we allow randomized policies. When the design decision is made, we are uncertain of some of the parameters governing the MDP, but we assume a distribution for these stochastic parameters is known. We focus on transient MDPs with a finite number of states and actions. We formulate, analyze and solve a two-stage stochastic integer program that yields an optimal design. A simple example threads its way through the paper to illustrate the development. The paper concludes with a larger application involving optimal design of malaria intervention strategies in Nigeria.