arXiv subjects

Edwin K. P. Chong

Publications and source records attributed to Edwin K. P. Chong.

At least 37 records · Page 2Linked to original sources

arXiv2018

Improved Bounds for the Greedy Strategy in Optimization Problems with Curvature

Consider the problem of choosing a set of actions to optimize an objective function that is a real-valued polymatroid function subject to matroid constraints. The greedy strategy provides an approximate solution to the optimization problem, and it is known to satisfy some performance bounds in terms of the total curvature. The total curvature depends on the value of objective function on sets outside the constraint matroid. If we are given a function defined only on the matroid, the problem still makes sense, but the existing bounds involving the total curvature do not apply. This is puzzling: If the optimization problem is perfectly well defined, why should the bounds no longer apply? This motivates an alternative formulation of such bounding techniques. The first question that comes to mind is whether it is possible to extend a polymatroid function defined on a matroid to one on the entire power set. This was recently shown to be negative in general. Here, we provide necessary and sufficient conditions for the existence of an \emph{incremental} extension of a polymatroid function defined on the uniform matroid of rank $k$ to one defined on the uniform matroid of rank $k+1$, together with an algorithm for constructing the extension. Whenever a polymatroid objective function defined on a matroid can be extended to the entire power set, the greedy approximation bounds involving the total curvature of the extension apply. However, these bounds still depend on sets outside the constraint matroid. Motivated by this, we define a new notion of curvature called \emph{partial curvature}, involving only sets in the matroid. We derive necessary and sufficient conditions for an extension of the function to have a total curvature that is equal to the partial curvature. Moreover, we prove that the bounds in terms of the partial curvature are in general improved over the previous ones......

math.OC↗

arXiv2018

A Modified Approximate Dynamic Programming Algorithm for Community-level Food Security Following Disasters

In the aftermath of an extreme natural hazard, community residents must have access to functioning food retailers to maintain food security. Food security is dependent on supporting critical infrastructure systems, including electricity, potable water, and transportation. An understanding of the response of such interdependent networks and the process of post-disaster recovery is the cornerstone of an efficient emergency management plan. In this study, the interconnectedness among different critical facilities, such as electrical power networks, water networks, highway bridges, and food retailers, is modeled. The study considers various sources of uncertainty and complexity in the recovery process of a community to capture the stochastic behavior of the spatially distributed infrastructure systems. The study utilizes an approximate dynamic programming (ADP) framework to allocate resources to restore infrastructure components efficiently. The proposed ADP scheme enables us to identify near-optimal restoration decisions at the community level. Furthermore, we employ a simulated annealing (SA) algorithm to complement the proposed ADP framework and to identify near-optimal actions accurately. In the sequel, we use the City of Gilroy, California, USA to illustrate the applicability of the proposed methodology following a severe earthquake. The approach can be implemented efficiently to identify practical policy interventions to hasten recovery of food systems and to reduce adverse food-insecurity impacts for other hazards and communities.

cs.CE↗

arXiv2018

Controlled Tracking in Urban Terrain: Closing the Loop

We investigate the challenging problem of integrating detection, signal processing, target tracking, and adaptive waveform scheduling with lookahead in urban terrain. We propose a closed-loop active sensing system to address this problem by exploiting three distinct levels of diversity: (1) spatial diversity through the use of coordinated multistatic radars; (2) waveform diversity by adaptively scheduling the transmitted waveform; and (3) motion model diversity by using a bank of parallel filters matched to different motion models. Specifically, at every radar scan, the waveform that yields the minimum trace of the one-step-ahead error covariance matrix is transmitted; the received signal goes through a matched-filter, and curve fitting is used to extract range and range-rate measurements that feed the LMIPDA-VSIMM algorithm for data association and filtering. Monte Carlo simulations demonstrate the effectiveness of the proposed system in an urban scenario contaminated by dense and uneven clutter, strong multipath, and limited line-of-sight.

eess.SP↗

arXiv2018

Near-optimal planning using approximate dynamic programming to enhance post-hazard community resilience management

The lack of a comprehensive decision-making approach at the community level is an important problem that warrants immediate attention. Network-level decision-making algorithms need to solve large-scale optimization problems that pose computational challenges. The complexity of the optimization problems increases when various sources of uncertainty are considered. This research introduces a sequential discrete optimization approach, as a decision-making framework at the community level for recovery management. The proposed mathematical approach leverages approximate dynamic programming along with heuristics for the determination of recovery actions. Our methodology overcomes the curse of dimensionality and manages multi-state, large-scale infrastructure systems following disasters. We also provide computational results showing that our methodology not only incorporates recovery policies of responsible public and private entities within the community but also substantially enhances the performance of their underlying strategies with limited resources. The methodology can be implemented efficiently to identify near-optimal recovery decisions following a severe earthquake based on multiple objectives for an electrical power network of a testbed community coarsely modeled after Gilroy, California, United States. The proposed optimization method supports risk-informed community decision makers within chaotic post-hazard circumstances.

math.OC↗

arXiv2018

Solving Markov decision processes for network-level post-hazard recovery via simulation optimization and rollout

Computation of optimal recovery decisions for community resilience assurance post-hazard is a combinatorial decision-making problem under uncertainty. It involves solving a large-scale optimization problem, which is significantly aggravated by the introduction of uncertainty. In this paper, we draw upon established tools from multiple research communities to provide an effective solution to this challenging problem. We provide a stochastic model of damage to the water network (WN) within a testbed community following a severe earthquake and compute near-optimal recovery actions for restoration of the water network. We formulate this stochastic decision-making problem as a Markov Decision Process (MDP), and solve it using a popular class of heuristic algorithms known as rollout. A simulation-based representation of MDPs is utilized in conjunction with rollout and the Optimal Computing Budget Allocation (OCBA) algorithm to address the resulting stochastic simulation optimization problem. Our method employs non-myopic planning with efficient use of simulation budget. We show, through simulation results, that rollout fused with OCBA performs competitively with respect to rollout with total equal allocation (TEA) at a meagre simulation budget of 5-10% of rollout with TEA, which is a crucial step towards addressing large-scale community recovery problems following natural disasters.

math.OC↗

arXiv2018

Stock Price Prediction using Principle Components

The literature provides strong evidence that stock prices can be predicted from past price data. Principal component analysis (PCA) is a widely used mathematical technique for dimensionality reduction and analysis of data by identifying a small number of principal components to explain the variation found in a data set. In this paper, we describe a general method for stock price prediction using covariance information, in terms of a dimension reduction operation based on principle component analysis. Projecting the noisy observation onto a principle subspace leads to a well-conditioned problem. We illustrate our method on daily stock price values for five companies in different industries. We investigate the results based on mean squared error and directional change statistic of prediction, as measures of performance, and volatility of prediction as a measure of risk.

q-fin.MF↗

arXiv2017

Performance Bounds for Nash Equilibria in Submodular Utility Systems with User Groups

In this paper, we consider variations of the utility system considered by Vetta, in which users are grouped together. Our aim is to establish how grouping and cooperation among users affect performance bounds. We consider two types of grouping. The first type is from \cite{Zhang2014}, where each user belongs to a group of users having social ties with it. For this type of utility system, each user's strategy maximizes its social group utility function, giving rise to the notion of \emph{social-aware Nash equilibrium}. We prove that this social utility system yields to the bounding results of Vetta for non-cooperative system, thus establishing provable performance guarantees for the social-aware Nash equilibrium. For the second type of grouping, the set of users is partitioned into $l$ disjoint groups, where the users within a group cooperate to maximize their group utility function, giving rise to the notion of \emph{group Nash equilibrium}. In this case, each group can be viewed as a new user with vector-valued actions, and a 1/2 bound for the performance of group Nash equilibrium follows from the result of Vetta. But as we show tighter bounds involving curvature can be established. By defining the group curvature $c_{k_i}$ associated with group $i$ with $k_i$ users, we show that if the social utility function is nondecreasing and submodular, then any group Nash equilibrium achieves at least $1/(1+\max_{1\leq i\leq l}c_{k_i})$ of the optimal social utility, which is tighter than that for the case without grouping. As a special case, if each user has the same action space, then we have that any group Nash equilibrium achieves at least $1/(1+c_{k^*})$ of the optimal social utility, where $k^*$ is the least number of users among the $l$ groups. Finally, we present an example of a utility system for database assisted spectrum access to illustrate our results.

math.OC↗

arXiv2017

Ranking and Selection as Stochastic Control

Under a Bayesian framework, we formulate the fully sequential sampling and selection decision in statistical ranking and selection as a stochastic control problem, and derive the associated Bellman equation. Using value function approximation, we derive an approximately optimal allocation policy. We show that this policy is not only computationally efficient but also possesses both one-step-ahead and asymptotic optimality for independent normal sampling distributions. Moreover, the proposed allocation policy is easily generalizable in the approximate dynamic programming paradigm.

cs.LG↗

arXiv2017

Performance Bounds with Curvature for Batched Greedy Optimization

The batched greedy strategy is an approximation algorithm to maximize a set function subject to a matroid constraint. Starting with the empty set, the batched greedy strategy iteratively adds to the current solution set a batch of elements that results in the largest gain in the objective function while satisfying the matroid constraints. In this paper, we develop bounds on the performance of the batched greedy strategy relative to the optimal strategy in terms of a parameter called the total batched curvature. We show that when the objective function is a polymatroid set function, the batched greedy strategy satisfies a harmonic bound for a general matroid constraint and an exponential bound for a uniform matroid constraint, both in terms of the total batched curvature. We also study the behavior of the bounds as functions of the batch size. Specifically, we prove that the harmonic bound for a general matroid is nondecreasing in the batch size and the exponential bound for a uniform matroid is nondecreasing in the batch size under the condition that the batch size divides the rank of the uniform matroid. Finally, we illustrate our results by considering a task scheduling problem and an adaptive sensing problem.

math.OC↗

arXiv2017

Polynomial-Time Methods to Solve Unimodular Quadratic Programs With Performance Guarantees

We develop polynomial-time heuristic methods to solve unimodular quadratic programs (UQPs) approximately, which are known to be NP-hard. In the UQP framework, we maximize a quadratic function of a vector of complex variables with unit modulus. Several problems in active sensing and wireless communication applications boil down to UQP. With this motivation, we present three new heuristic methods with polynomial-time complexity to solve the UQP approximately. The first method is called dominant-eigenvector-matching; here the solution is picked that matches the complex arguments of the dominant eigenvector of the Hermitian matrix in the UQP formulation. We also provide a performance guarantee for this method. The second method, a greedy strategy, is shown to provide a performance guarantee of (1-1/e) with respect to the optimal objective value given that the objective function possesses a property called string submodularity. The third heuristic method is called row-swap greedy strategy, which is an extension to the greedy strategy and utilizes certain properties of the UQP to provide a better performance than the greedy strategy at the expense of an increase in computational complexity. We present numerical results to demonstrate the performance of these heuristic methods, and also compare the performance of these methods against a standard heuristic method called semidefinite relaxation.

math.OC↗

arXiv2016

Performance Bounds for the $k$-Batch Greedy Strategy in Optimization Problems with Curvature

The $k$-batch greedy strategy is an approximate algorithm to solve optimization problems where the optimal solution is hard to obtain. Starting with the empty set, the $k$-batch greedy strategy adds a batch of $k$ elements to the current solution set with the largest gain in the objective function while satisfying the constraints. In this paper, we bound the performance of the $k$-batch greedy strategy with respect to the optimal strategy by defining the total curvature $α_k$. We show that when the objective function is nondecreasing and submodular, the $k$-batch greedy strategy satisfies a harmonic bound $1/(1+α_k)$ for a general matroid constraint and an exponential bound $\left(1-(1-α_k/{t})^t\right)/α_k$ for a uniform matroid constraint, where $k$ divides the cardinality of the maximal set in the general matroid, $t=K/k$ is an integer, and $K$ is the rank of the uniform matroid. We also compare the performance of the $k$-batch greedy strategy with that of the $k_1$-batch greedy strategy when $k_1$ divides $k$. Specifically, we prove that when the objective function is nondecreasing and submodular, the $k$-batch greedy strategy has better harmonic and exponential bounds in terms of the total curvature. Finally, we illustrate our results by considering a task-assignment problem.

math.OC↗

arXiv2016

Bounding the Greedy Strategy in Finite-Horizon String Optimization

We consider an optimization problem where the decision variable is a string of bounded length. For some time there has been an interest in bounding the performance of the greedy strategy for this problem. Here, we provide weakened sufficient conditions for the greedy strategy to be bounded by a factor of $(1-(1-1/K)^K)$, where $K$ is the optimization horizon length. Specifically, we introduce the notions of $K$-submodularity and $K$-GO-concavity, which together are sufficient for this bound to hold. By introducing a notion of \emph{curvature} $η\in(0,1]$, we prove an even tighter bound with the factor $(1/η)(1-e^{-η})$. Finally, we illustrate the strength of our results by considering two example applications. We show that our results provide weaker conditions on parameter values in these applications than in previous results.

math.OC↗

arXiv2015

Subspace selection for projection maximization with matroid constraints

Suppose that there is a ground set which consists of a large number of vectors in a Hilbert space. Consider the problem of selecting a subset of the ground set such that the projection of a vector of interest onto the subspace spanned by the vectors in the chosen subset reaches the maximum norm. This problem is generally NP-hard, and alternative approximation algorithms such as forward regression and orthogonal matching pursuit have been proposed as heuristic approaches. In this paper, we investigate bounds on the performance of these algorithms by introducing the notions of elemental curvatures. More specifically, we derive lower bounds, as functions of these elemental curvatures, for performance of the aforementioned algorithms with respect to that of the optimal solution under uniform and non-uniform matroid constraints, respectively. We show that if the elements in the ground set are mutually orthogonal, then these algorithms are optimal when the matroid is uniform and they achieve at least $1/2$-approximations of the optimal solution when the matroid is non-uniform.

cs.IT↗

arXiv2015

Intersection Information based on Common Randomness

The introduction of the partial information decomposition generated a flurry of proposals for defining an intersection information that quantifies how much of "the same information" two or more random variables specify about a target random variable. As of yet, none is wholly satisfactory. A palatable measure of intersection information would provide a principled way to quantify slippery concepts, such as synergy. Here, we introduce an intersection information measure based on the Gács-Körner common random variable that is the first to satisfy the coveted target monotonicity property. Our measure is imperfect, too, and we suggest directions for improvement.

cs.IT↗

arXiv2015

String Submodular Functions with Curvature Constraints

The problem of objectively choosing a string of actions to optimize an objective function that is string submodular has been considered in [1]. There it is shown that the greedy strategy, consisting of a string of actions that only locally maximizes the step-wise gain in the objective function achieves at least a (1-e^{-1})-approximation to the optimal strategy. This paper improves this approximation by introducing additional constraints on curvatures, namely, total backward curvature, total forward curvature, and elemental forward curvature. We show that if the objective function has total backward curvature σ, then the greedy strategy achieves at least a \frac{1}σ(1-e^{-σ})-approximation of the optimal strategy. If the objective function has total forward curvature ε, then the greedy strategy achieves at least a (1-ε)-approximation of the optimal strategy. Moreover, we consider a generalization of the diminishing-return property by defining the elemental forward curvature. We also consider the problem of maximizing the objective function subject to general a string-matroid constraint. We investigate an applications of string submodular functions with curvature constraints.

cs.DS↗

arXiv2015

Robust Decentralized Detection and Social Learning in Tandem Networks

We study a tandem of agents who make decisions about an underlying binary hypothesis, where the distribution of the agent observations under each hypothesis comes from an uncertainty class. We investigate both decentralized detection rules, where agents collaborate to minimize the error probability of the final agent, and social learning rules, where each agent minimizes its own local minimax error probability. We then extend our results to the infinite tandem network, and derive necessary and sufficient conditions on the uncertainty classes for the minimax error probability to converge to zero when agents know their positions in the tandem. On the other hand, when agents do not know their positions in the network, we study the cases where agents collaborate to minimize the asymptotic minimax error probability, and where agents seek to minimize their worst-case minimax error probability (over all possible positions in the tandem). We show that asymptotic learning of the true hypothesis is no longer possible in these cases, and derive characterizations for the minimax error performance.

cs.IT↗

arXiv2014

Guaranteed Bounds for General Approximate Dynamic Programming

In this paper, we will develop a systematic approach to deriving guaranteed bounds for approximate dynamic programming (ADP) schemes in optimal control problems. Our approach is inspired by our recent results on bounding the performance of greedy strategies in optimization of string-submodular functions over a finite horizon. The approach is to derive a string-submodular optimization problem, for which the optimal strategy is the optimal control solution and the greedy strategy is the ADP solution. Using this approach, we show that any ADP solution achieves a performance that is at least a factor of $β$ of the performance of the optimal control solution, which satisfies Bellman's optimality principle. The factor $β$ depends on the specific ADP scheme, as we will explicitly characterize. To illustrate the applicability of our bounding technique, we present examples of ADP schemes, including the popular rollout method.

math.OC↗

arXiv2013

Hypothesis Testing in Feedforward Networks with Broadcast Failures

Consider a countably infinite set of nodes, which sequentially make decisions between two given hypotheses. Each node takes a measurement of the underlying truth, observes the decisions from some immediate predecessors, and makes a decision between the given hypotheses. We consider two classes of broadcast failures: 1) each node broadcasts a decision to the other nodes, subject to random erasure in the form of a binary erasure channel; 2) each node broadcasts a randomly flipped decision to the other nodes in the form of a binary symmetric channel. We are interested in whether there exists a decision strategy consisting of a sequence of likelihood ratio tests such that the node decisions converge in probability to the underlying truth. In both cases, we show that if each node only learns from a bounded number of immediate predecessors, then there does not exist a decision strategy such that the decisions converge in probability to the underlying truth. However, in case 1, we show that if each node learns from an unboundedly growing number of predecessors, then the decisions converge in probability to the underlying truth, even when the erasure probabilities converge to 1. We also derive the convergence rate of the error probability. In case 2, we show that if each node learns from all of its previous predecessors, then the decisions converge in probability to the underlying truth when the flipping probabilities of the binary symmetric channels are bounded away from 1/2. In the case where the flipping probabilities converge to 1/2, we derive a necessary condition on the convergence rate of the flipping probabilities such that the decisions still converge to the underlying truth. We also explicitly characterize the relationship between the convergence rate of the error probability and the convergence rate of the flipping probabilities.

cs.IT↗