Results 1 
7 of
7
Nearoptimal sensor placements in gaussian processes
 In ICML
, 2005
"... When monitoring spatial phenomena, which can often be modeled as Gaussian processes (GPs), choosing sensor locations is a fundamental task. There are several common strategies to address this task, for example, geometry or disk models, placing sensors at the points of highest entropy (variance) in t ..."
Abstract

Cited by 174 (27 self)
 Add to MetaCart
When monitoring spatial phenomena, which can often be modeled as Gaussian processes (GPs), choosing sensor locations is a fundamental task. There are several common strategies to address this task, for example, geometry or disk models, placing sensors at the points of highest entropy (variance) in the GP model, and A, D, or Eoptimal design. In this paper, we tackle the combinatorial optimization problem of maximizing the mutual information between the chosen locations and the locations which are not selected. We prove that the problem of finding the configuration that maximizes mutual information is NPcomplete. To address this issue, we describe a polynomialtime approximation that is within (1 − 1/e) of the optimum by exploiting the submodularity of mutual information. We also show how submodularity can be used to obtain online bounds, and design branch and bound search procedures. We then extend our algorithm to exploit lazy evaluations and local structure in the GP, yielding significant speedups. We also extend our approach to find placements which are robust against node failures and uncertainties in the model. These extensions are again associated with rigorous theoretical approximation guarantees, exploiting the submodularity of the objective function. We demonstrate the advantages of our approach towards optimizing mutual information in a very extensive empirical study on two realworld data sets.
Sensor selection via convex optimization
 IEEE Transactions on Signal Processing
, 2009
"... Abstract—We consider the problem of choosing a set of sensor measurements, from a set of possible or potential sensor measurements, that minimizes the error in estimating some parameters. Solving this problem by evaluating the performance for each of the possible choices of sensor measurements is no ..."
Abstract

Cited by 22 (2 self)
 Add to MetaCart
Abstract—We consider the problem of choosing a set of sensor measurements, from a set of possible or potential sensor measurements, that minimizes the error in estimating some parameters. Solving this problem by evaluating the performance for each of the possible choices of sensor measurements is not practical unless and are small. In this paper, we describe a heuristic, based on convex optimization, for approximately solving this problem. Our heuristic gives a subset selection as well as a bound on the best performance that can be achieved by any selection of sensor measurements. There is no guarantee that the gap between the performance of the chosen subset and the performance bound is always small; but numerical experiments suggest that the gap is small in many cases. Our heuristic method requires on the order of operations; for 1000 possible sensors, we can carry out sensor selection in a few seconds on a 2GHz personal computer. Index Terms—Convex optimization, experiment design, sensor selection. I.
An algorithm for constructing orthogonal and nearlyorthogonal arrays with mixed levels and small runs
 Technometrics
, 2002
"... Orthogonal arrays are used widely in manufacturing and hightechnology industries for quality and productivity improvement experiments. For reasons of run size economy or � exibility, nearlyorthogonal arrays are also used. The construction of orthogonal or nearlyorthogonal arrays can be quite chal ..."
Abstract

Cited by 9 (3 self)
 Add to MetaCart
Orthogonal arrays are used widely in manufacturing and hightechnology industries for quality and productivity improvement experiments. For reasons of run size economy or � exibility, nearlyorthogonal arrays are also used. The construction of orthogonal or nearlyorthogonal arrays can be quite challenging. Most existing methods are complex and produce limited types of arrays. This article describes a simple and effective algorithm for constructing mixedlevel orthogonal and nearlyorthogonal arrays that can construct a variety of smallrun designs with good statistical properties ef � ciently. KEY WORDS: Doptimality; Exchange algorithm; Interchange algorithm; J 2optimality. 1.
Design Of Spatial Experiments: Model Fitting And Prediction
"... The main objective of the paper is to describe and develop model oriented methods and algorithms for the design of spatial experiments. Unlike many other publications in this area, the approach proposed here is essentially based on the ideas of convex design theory. 1. ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
The main objective of the paper is to describe and develop model oriented methods and algorithms for the design of spatial experiments. Unlike many other publications in this area, the approach proposed here is essentially based on the ideas of convex design theory. 1.
Testing of Analog Systems Using Behavioral Models and Optimal Experimental Design Techniques
 Proc. IEEE ICCAD
, 1994
"... This paper describesa new CAD algorithm which performsautomatic test pattern generation (ATPG) for a general class of analog systems, namely those circuits which can be efficiently modeled as an additive combination of userdefined basis functions. The algorithm is based on the statistical technique ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
This paper describesa new CAD algorithm which performsautomatic test pattern generation (ATPG) for a general class of analog systems, namely those circuits which can be efficiently modeled as an additive combination of userdefined basis functions. The algorithm is based on the statistical technique of Ioptimal experimental design, in which test vectors are chosen to be maximally independent so that circuit performance will be characterized as accurately as possible in the presence of measurement noise and model inaccuracies. This technique allows analog systems to be characterized more accurately and more efficiently, thereby significantly reducing system test time and hence total manufacturing cost. 1 Introduction The complexity of electronic systems being designed today is increasing in many dimensions: on one hand, the number of components is growing constantly; on the other, several radically different functions must be integrated. For example, in the exploding personal communi...
SUBSET SELECTION AND OPTIMIZATION FOR SELECTING BINOMIAL SYSTEMS APPLIED TO SUPERSATURATED DESIGN GENERATION
"... The problem of finding the binomial population with the highest success probability is considered when the number of binomial populations is large. A new rigorous indifference zone subset selection procedure for binomial populations is proposed with the proof of the corresponding least favorable con ..."
Abstract
 Add to MetaCart
The problem of finding the binomial population with the highest success probability is considered when the number of binomial populations is large. A new rigorous indifference zone subset selection procedure for binomial populations is proposed with the proof of the corresponding least favorable configuration. For cases involving large numbers of binomial populations, a simulation optimization method combining the proposed subset selection procedure with an elitist Genetic Algorithm (GA) is proposed to find the highestmean solution. Convergence of the proposed GA frame work are established under general assumptions. The problem of deriving supersaturated screening designs is described and used to illustrate the application of all methods. Computational comparisons are also presented for the problem of generating supersaturated experimental designs. 1
unknown title
, 2005
"... www.elsevier.com/locate/apthermeng Multiperiod steam turbine network optimisation. Part I: Simulation based regression models and an evolutionary algorithm for finding Doptimal designs ..."
Abstract
 Add to MetaCart
www.elsevier.com/locate/apthermeng Multiperiod steam turbine network optimisation. Part I: Simulation based regression models and an evolutionary algorithm for finding Doptimal designs