Bayesian selling problem with partial information

We introduce an optimal stopping problem for selling an asset when the fixed but unknown distribution of successive offers is from one of n possible distributions. The initial probabilities as to which is the true distribution are given and updated in a Bayesian manner as the successive offers are observed. After receiving an offer, the seller has to decide whether to accept the offer or continue to observe the next offer. Each time an offer is observed a fixed cost is incurred. We consider both the cases where recalling a past offer is allowed and where it is not allowed. For each case, a dynamic programming model and some heuristic policies are presented. Using simulation, the performances of the heuristic methods are evaluated and upper bounds on the optimal expected return are obtained. © 2013 Wiley Periodicals, Inc. Naval Research Logistics, 2013     

Bayesian sampling plans for exponential distributions with interval censored samples

This article studies the problem of designing Bayesian sampling plans (BSP) with interval censored samples. First, an algorithm for deriving the conventional BSP is proposed. The BSP is shown to possess some monotonicity. Based on the BSP and using the property of monotonicity, a new sampling plan modified by the curtailment procedure is proposed. The resulting curtailed Bayesian sampling plan (CBSP) can reduce the duration time of life test experiment, and it is optimal in the sense that its associated Bayes risk is smaller than the Bayes risk of the BSP if the cost of the duration time of life test experiment is considered. A numerical example to compute the Bayes risks of BSP and CBSP and related quantities is given. Also, a Monte Carlo simulation study is performed to illustrate the performance of the CBSP compared with the BSP. The simulation results demonstrate that our proposed CBSP has better performance because it has smaller risk. The CBSP is recommended. © 2015 Wiley Periodicals, Inc. Naval Research Logistics 62: 604–616, 2015     

Sequential learning versus no learning in Bayesian regression models

We consider two regression models: linear and logistic. The dependent variable is observed periodically and in each period a Bayesian formulation is used to generate updated forecasts of the dependent variable as new data is observed. One would expect that including new data in the Bayesian updates results in improved forecasts over not including the new data. Our findings indicate that this is not always true. We show there exists a subset of the independent variable space that we call the “region of no learning.” If the independent variable values for a given period in the future are in this region, then the forecast does not change with any new data. Moreover, if the independent variable values are in a neighborhood of the region of no learning, then there may be little benefit to wait for the new data and update the forecast. We propose a statistical approach to characterize this neighborhood which we call the “region of little learning.” Our results provide insights into the trade‐offs that exist in situations when the decision maker has an incentive to make an early decision based on an early forecast versus waiting to make a later decision based on an updated forecast. © 2014 Wiley Periodicals, Inc. Naval Research Logistics 61: 532–548, 2014     

Asymmetric information and third-party intervention in civil wars

I study a two-period model of conflict with two combatants and a third party who is an ally of one of the combatants. The third party is fully informed about the type of her ally but not about the type of her ally’s enemy. In a signaling game, I find that if the third party is unable to give a sufficiently high assistance to her ally, then there exists a unique separating equilibrium in which the third party’s expected intervention causes her ally’s enemy to exert more effort than in the absence of third-party intervention; this worsens the conflict.     

机动目标跟踪的非线性算法

卡尔曼滤波器对线性高斯滤波问题能提供最优解, 而对目标运动模型、观测方程等要求的非线性就不再适合,提出了一种机动目标自适应非线性粒子滤波算法-" 粒子滤波器"(Particle Filters PF)法, 这种方法不受线性化误差和高斯噪声假定的限制,适用于任何状态转换或测量模型, 分析比较了粒子滤波(PF)与扩展卡尔曼滤波算法(EKF) 的滤波精度、运算量等方面指标.给出了基于典型非线性模型的算法仿真, 仿真结果表明粒子滤波新方法优于EKF对机动目标跟踪.     

高超声速滑翔再入飞行器的可达区快速预测

防御方对来袭滑翔再入飞行器进行可达区的预测存在先验信息量不足且时效性要求高等难题.为此提出一种基于最优化飞行假设的可达区快速预测方法:仅需已知目标当前位置、速度与最大升阻比(可基于雷达探测数据通过实时弹道估计获得),基于平衡滑翔假设和最大横程的埃格斯解分别获得目标最大纵程和横程终点坐标,在经纬度二维平面内,可达区即可近...