Threshold Accepting Research Articles

Search for good policies in a single-warehouse, multi-retailer system by particle swarm optimisation

An optimal control of sufficiently realistic single-warehouse, multi-retailer systems (SWMR-systems) requires the simulation optimisation approach. After a brief discussion of modelling and simulation aspects for such systems, we present a simulator providing a sufficiently large set of options to customize it to realistic problem situations. The paper then focuses on the applicability of optimisation algorithms and introduces the methods particle swarm optimisation and threshold accepting. We show that both approaches yield good optimisation results for several examples of a 5-retailer system investigated according to their reordering strategy and transportation resources.

Optimized -type designs on flexible regions

The concept of a flexible region describes an infinite variety of symmetrical shapes to enclose a particular region of interest within a space. In experimental design, the properties of a function on the region of interest are analyzed based on a set of design points. The choice of design points can be made based on some discrepancy criterion. The generation of design points on a flexible region is investigated. A recently proposed discrepancy measure, the central composite discrepancy, is used for this purpose. The optimization heuristic Threshold Accepting is applied to generate low-discrepancy U -type designs. The proposed algorithm is capable of constructing optimal U -type designs under various flexible experimental regions in two or more dimensions. The illustrative results for the two-dimensional case indicate that by using an optimization heuristic in combination with an appropriate discrepancy measure produces high-quality experimental designs on flexible regions.

Applying the threshold accepting metaheuristic to curriculum based course timetabling

The article presents a study of local search algorithms for timetabling problems, with the particular goal of providing a contribution to competition track 3 of the International Timetabling Competition 2007 (ITC 2007). In this track, a formulation of a curriculum based course timetabling has been published, and novel benchmark instances have been presented that allow the comparison of optimization approaches. Our heuristic local search procedure is based on the principles of Threshold Accepting, overcoming local optima by a deterministic acceptance of inferior solutions throughout the search runs. A stochastic neighborhood is proposed and implemented, randomly removing and reassigning events from the current solution. The overall concept has been incrementally obtained from a series of experiments, which we describe in each (sub)section of the paper. In conclusions, we successfully derived a potential candidate solution approach for the finals of track 3 of the ITC 2007, held in August 2008 in Montréal, Canada.

Threshold accepting framework for discrete and continuous search spaces

In this paper, a threshold accepting multi-agents (TA-MA) framework is proposed to optimise discrete and continuous problems. The TA-MA framework uses two agents, in each one a threshold accepting (TA) metaheuristic is started with different initial parameters. The cooperation between these agents is assured by an adaptive memory named taboo central memory (TCM). Computational experiments, on the single machine total weighted tardiness (SMTWT) problems and on two engineering design optimisation problems (the supported I-beam and the machine tool spindle), show that the proposed TA-MA algorithm can obtain the optimal solution or much better than the best known one for all standard benchmark problems instances from the literature.

Differential evolution and threshold accepting hybrid algorithm for unconstrained optimisation

This paper develops a hybrid global optimisation metaheuristic methodology for solving unconstrained optimisation problems. The hybrid, to be called DETA, comprises two global optimisation algorithms viz. differential evolution (DE) and threshold accepting (TA) in tandem. While working with DE on benchmark problems, we noticed that it slows down before convergence is achieved. After analysing the possible reason for this shortcoming, we propose DETA to address it. DETA works in two phases: Phase 1 implements the original DE with relaxed convergence criterion. Then, a switch over is made from Phase 1 to Phase 2, where TA is used to quickly guide the search to global optimum. Performance of DETA is compared with that of DE on 26 unconstrained benchmark problems. The results obtained indicate that DETA hybrid is much superior to DE in terms of speed for the same level of accuracy.

Heuristic optimization of RC bridge piers with rectangular hollow sections

This paper deals with the economic optimization of reinforced concrete (RC) bridge piers with hollow rectangular sections and describes the efficiency of three heuristic algorithms: two new variants of the ant colony optimization (ACO) algorithm, the genetic algorithm (GA) and the threshold acceptance (TA) algorithm. The GA and TA are used for comparison with the new ACO algorithms. The total number of variables is 95. All variables are discrete in this analysis. The calibration of the new ACO algorithm recommended a 250-member ant population and 100 stages. The best solution costs 69,467 euros, which means savings of about 33% as compared to experience-based design. Finally, results indicate that the new ACO algorithms are potentially useful for optimizing the costs of real RC structures.

Distributed optimisation of a portfolio’s Omega

We investigate portfolio selection with an alternative objective function in a distributed computing environment. More specifically, we optimise a portfolio’s ‘Omega’ which is the ratio of two partial moments of the portfolio’s return distribution. Since finding optimal portfolios under such a performance measure and realistic constraints is a non-convex problem, we suggest to solve the problem with a heuristic method called Threshold Accepting (TA). TA is a very flexible technique as it requires no simplifications of the problem and allows for a straightforward implementation of all kinds of constraints. Applying the algorithm to actual data, we find that TA is well-capable of solving this type of problem. Furthermore, we show that the computations can easily be distributed which leads to considerable speedups.

Least median of squares estimation by optimization heuristics with an application to the CAPM and a multi-factor model

For estimating the parameters of models for financial market data, the use of robust techniques is of particular interest. Conditional forecasts, based on the capital asset pricing model, and a factor model are considered. It is proposed to consider least median of squares estimators as one possible alternative to ordinary least squares. Given the complexity of the objective function for the least median of squares estimator, the estimates are obtained by means of optimization heuristics. The performance of two heuristics is compared, namely differential evolution and threshold accepting. It is shown that these methods are well suited to obtain least median of squares estimators for real world problems. Furthermore, it is analyzed to what extent parameter estimates and conditional forecasts differ between the two estimators. The empirical analysis considers daily and monthly data on some stocks from the Dow Jones Industrial Average Index.

Improving Quality

In 1967, the first multicenter randomized controlled trial of treatment for hypertension was published by the newly created Veterans Administration Cooperative Studies Program, demonstrating the remarkable effectiveness of drug treatment for what would now be considered severe hypertension (average diastolic pressure 115 to 129 mm Hg).1 During the ensuing 4 decades, there have been remarkable strides made to improve care for this very common condition, including development of effective new drugs and acceptance of lower blood pressure thresholds and targets. It is more than fitting that the Department of Veterans Affairs (VA) should remain a leading contributor to these efforts. Article see p 392 Despite these advances, however, efforts to improve awareness of high blood pressure and achievement of treatment goals remains suboptimal from a national perspective, particularly among blacks and other ethnic minorities.2 Increasingly it is being acknowledged that additional progress depends far more on improving systems of care than further advances in pharmacology. Because they are frequently characterized by defined populations, substantial administrative resources, infrastructure such as information systems, an emphasis on primary care, and coordinated benefits, large integrated health systems have been most successful in implementing improvements of care for chronic conditions. In this issue of Circulation: Cardiovascular Quality and Outcomes , Choma et al3 describe the results of efforts to improve the quality of monitoring and treatment of hypertension in 1 of the 21 regional networks (Veterans Integrated Service Networks [VISNs]) that comprise the VA health care system. To be appreciated, the results reported by Choma et al must be viewed in the larger context of the overarching Veterans Health Administration (VHA). The largest integrated healthcare system in the United States, VHA has more than 200 000 employees (including 17 000 physicians) and …

A well-scalable metaheuristic for the fleet size and mix vehicle routing problem with time windows

This paper presents an efficient and well-scalable metaheuristic for fleet size and mix vehicle routing with time windows. The suggested solution method combines the strengths of well-known threshold accepting and guided local search metaheuristics to guide a set of four local search heuristics. The computational tests were done using the benchmarks of [Liu, F.-H., & Shen, S.-Y. (1999). The fleet size and mix vehicle routing problem with time windows. Journal of the Operational Research Society, 50(7), 721–732] and 600 new benchmark problems suggested in this paper. The results indicate that the suggested method is competitive and scales almost linearly up to instances with 1000 customers.

Constructing Long/Short Portfolios with the Omega Ratio

We construct portfolios with an alternative selection criterion, the Omega function, which can be expressed as the ratio of two partial moments of the returns distribution. Finding Omega-optimal portfolios, in particular under realistic constraints like cardinality restrictions, requires to solve non-convex optimisation problems. Since standard (gradient-based) optimisation methods fail here, we suggest to use a heuristic technique (Threshold Accepting). The main purpose of the paper is to investigate the empirical performance of the selected portfolios, especially the effects of allowing short positions. Many studies on portfolio optimisation assume that short sales are not allowed. This is despite the fact that theoretically, short positions can improve the risk-return characteristics of a portfolio, and practically, institutional investors can and do sell stocks short. We investigate whether removing the non-negativity constraint really improves out-of-sample portfolio performance under realistic assumptions, that is when optimal weights need to be estimated from the data, different transaction costs apply to long and short positions or short selling is restricted to specific assets.

Practical Remedial Design Optimization for Large Complex Plumes

Powerful simulation/optimization (S/O) models exist for designing groundwater well systems and pumping strategies. However, it can be challenging to use S/O modeling effectively for large, complex, and computationally intensive problems within project time and cost constraints. Here, we present a generic two-stage optimization procedure for making S/O modeling more practical. Application is illustrated for developing optimal transient 30-year pump-and-treat designs for Blaine Naval Ammunition Depot (NAD), Nebraska, and using an innovative hybrid advanced genetic algorithm with tabu search features (AGT). AGT includes standard genetic algorithm and tabu search features plus healing, elitism, threshold acceptance, and a new subset/subspace decomposition optimization. The screening stage simplifies the optimization problem, and selects desirable remediation wells from among many candidates. During this stage, computational effort is lessened by reducing the number of state variables needing evaluation, and the solution space dimensionality (including temporal dimensions). Subset/subspace decomposition optimization of steady flow rates is used to identify desirable sets of candidate wells. The transient optimization stage develops mathematically optimal time-varying pumping rates for well subsets identified by the screening stage. It also includes reoptimization using the original objective function plus goal programming to increase strategy robustness. Initializing the AGT with feasible solutions reduces computational effort. Within a short period the procedure developed optimal pump and treat system designs for NAD. The procedure yields better objective function values than trial and error. Because optimization causes tight constraints, the computed strategy is sensitive to changes in model parameters. Increasing strategy robustness via AGT and goal programming degrades the value of the initial objective function.

An hybrid simulated annealing threshold accepting algorithm for satisfiability problems using dynamically cooling schemes

For Satisfiability (SAT) Problem there is not a deterministic algorithm able to solve it in a polynomial time. Simulated Annealing (SA) and similar algorithms like Threshold Accepting (TA) are able...

Threshold accepting and Ant-colony optimization algorithms for scheduling m-machine flow shops with lot streaming

This paper addresses the problem of making sequencing and scheduling decisions for n jobs m machines flow shops under lot sizing environment. Lot streaming (lot sizing) is the process of creating sublots to move the completed portion of a production sublots to down stream machines. There is a scope for efficient algorithms for scheduling problems in m-machine flow shop with lot streaming. In recent years, much attention is given to heuristics and search techniques. On this concern this paper proposes Ant-colony optimization algorithm (ACO) and threshold accepting algorithm (TA) to evolve best sequence for makespan/total flow time criterion for m-machine flow shop involved with lot streaming and setup time. The following two algorithms are used to evaluate the performance of the proposed ACO and TA: (i) Baker's algorithm (BA), an optimal solution procedure for two-machine flow shop problem with lot streaming and makespan objective criterion and (ii) genetic and hybrid genetic algorithm for m-machine flow shop problem with lot streaming and makespan and total flow time criteria.

Estimating the effects of adjacency and green-up constraints on landowners of different sizes and spatial arrangements located in the southeastern U.S.

The maximum clearcut size and green-up period is important for land managers adhering to voluntary and regulatory guidelines. Therefore the impact of actual and hypothetical clearcut size restrictions is a concern for forest landowners who manage land and intend to practice forestry for profit. In this research, the effect of a 97.1 ha (240 ac) clearcut size constraint with a green-up period of 2-yrs is assessed for forest landowners with different forest land sizes, ownership patterns, and age class distributions. A meta heuristic which consists of threshold accepting, 1-opt tabu search, and 2-opt tabu search is used to develop spatially-constrained forest plans for 27 hypothetical forest landowners. These results are compared to a relaxed solution produced with linear programming, and statistical analyses are used to determine significant differences. The analysis provided enough evidence to suggest that two factors (size of ownership pattern and initial age class distribution), and one interaction factor (ownership size × initial age class distribution) are significant in explaining the differences in the percent reduction in forest plan value among the forests managed by the hypothetical forest landowners. From an absolute value reduction perspective, small-sized older forests were most affected. From a percent value point of view, one can conclude from this analysis that landowners with small-sized forests and young initial age class distributions will be significantly more affected by potential adjacency and green-up restrictions in the southeastern U.S. than other types of landowners.

Smooth Transition Autoregressive Models -- New Approaches to the Model Selection Problem

It has been shown in the literature that the task of estimating the parameters of nonlinear models may be tackled with optimization heuristics. Thus, we attempt to carry these intuitions over to the estimation procedure of smooth transition autoregressive (STAR, Terasvirta, 1994) models by introducing the following three stochastic optimization algorithms: Simulated Annealing, (Kirkpatrick, Gelatt, and Vecchi, 1983), Threshold Accepting (Dueck and Scheuer, 1990) and Differential Evolution (Storn and Price, 1995, 1997). Besides considering the performance of these heuristics in estimating STAR model parameters, our paper additionally picks up the problem of identifying redundant parameters which, according to our view, has not been addressed in a satisfactory way by now. The resulting findings of our simulation studies seem to argue for an implementation of heuristic approaches within the STAR modeling cycle. In particular for the case of STAR model specification, an application of these heuristics might offer valuable information to empirical researchers.

Software development cost estimation using wavelet neural networks

Software development has become an essential investment for many organizations. Software engineering practitioners have become more and more concerned about accurately predicting the cost and quality of software product under development. Accurate estimates are desired but no model has proved to be successful at effectively and consistently predicting software development cost. In this paper, we propose the use of wavelet neural network (WNN) to forecast the software development effort. We used two types of WNN with Morlet function and Gaussian function as transfer function and also proposed threshold acceptance training algorithm for wavelet neural network (TAWNN). The effectiveness of the WNN variants is compared with other techniques such as multilayer perceptron (MLP), radial basis function network (RBFN), multiple linear regression (MLR), dynamic evolving neuro-fuzzy inference system (DENFIS) and support vector machine (SVM) in terms of the error measure which is mean magnitude relative error (MMRE) obtained on Canadian financial (CF) dataset and IBM data processing services (IBMDPS) dataset. Based on the experiments conducted, it is observed that the WNN-Morlet for CF dataset and WNN-Gaussian for IBMDPS outperformed all the other techniques. Also, TAWNN outperformed all other techniques except WNN.

Distributed Optimisation of a Portfolio's Omega

We investigate portfolio selection with alternative objective functions in a distributed computing environment. In particular, we optimise a portfolio's 'Omega' which is the ratio of two partial moments of the returns distributions. Since finding optimal portfolios under such performance measures and realistic constraints leads to non-convex problems, we suggest to solve the problem with a heuristic method called Threshold Accepting (TA). TA is a very flexible technique as it requires no simplifications of the problem and allows for a straightforward implementation of all kinds of constraints. Applying this algorithm to actual data, we find that TA is well-adapted to optimisation problems of this type. Furthermore, we show that the computations can easily be distributed which leads to considerable speedups.

Threshold accepting trained principal component neural network and feature subset selection: Application to bankruptcy prediction in banks

This paper proposes an application of new principal component neural network (PCNN) architecture to bankruptcy prediction problem in commercial banks. Further, a new feature subset selection (FSS) algorithm is proposed. In this architecture, the hidden layer is completely replaced by what is referred to as a ‘principal component layer’. This layer consists of a few selected principal components that perform the function of hidden nodes. Moreover, this study proposes an algorithm based on the threshold accepting (TA) meta-heuristic to train the PCNN. The architecture reduces the number of weights by a great number as there are no formal connections between the input layer and the principal component layer. The efficacy of the algorithm is tested on the Spanish banks dataset and Turkish banks dataset. The results showed high generalization power of PCNN in the 10-fold cross-validation and also the feature subsets selected in each of the examples showed high discriminating power. PCNN is also compared with PCA-TANN and PCA-BPNN, which have PCA as the preprocessor and have one hidden layer each. Further comparisons are also made with TANN and BPNN. All these classifiers are compared with respect to the AUC (area under the receiver operating characteristic (ROC) curve) criterion. ROC curve is drawn for each classifier with sensitivity on the X-axis and one-specificity on the Y-axis. Based on the experiments conducted, it is inferred that the proposed PCNN hybrids outperformed other classifiers in terms of AUC. It is also observed that the proposed feature subset selection algorithm is very stable and powerful.

Design of reinforced concrete bridge frames by heuristic optimization

This paper deals with the economic optimization of reinforced concrete box frames used in road construction. It shows the efficiency of four heuristic algorithms applied to a problem of 50 design variables. Heuristic methods used are the random walk and the descent local search. The metaheuristic methods are the threshold accepting and the simulated annealing. The four methods have been applied to the same frame of 13 m of horizontal span. The comparison of the four heuristic algorithms leads to the conclusion that the proposed threshold accepting is more efficient, since it improves cost results of the random walk and descent local search by 7.5% and 1.4%, respectively, while improving deviation of random results of the simulated annealing. Finally, the inclusion of the deflections and fatigue limit states appears to be crucial, since their ignorance leads to 3.9% more economic but unsafe results.