Multiresolution and local search methods for optimizing visual tracking processes on GPU

Robust face alignment and tracking by combining local search and global fitting

Tabu Search is one of the local search methods used for mathematical optimization Metaheuristics search method. It was founded in 1986 by Fred W. Developed by Clover and in 1989 Formalized. Local (nearby) searches take a potential solution to a problem and its immediate neighbor Check countries (i.e., similar solutions except for very small details). Improved solution Diagnosis. Local search methods on plateaus where subdivisions or multiple solutions are equally applicable Tend to get entangled. Tabu Search is the local search by relaxing its basic rule Improves performance. First, any moves that get worse with each step Will be accepted if not upgraded (if the search is stuck in a strict local minimum). In addition, obstacles (hereinafter referred to as taboo) prevent the return to previously visited solutions Introduced in the category. The implementation of the tab search, the solutions visited or the user Uses memory systems that describe sets of rules provided. [2] A certain short If the possible solution within the period has been visited before or if it violates a rule, it is Will be referred to as "taboo" (blocked) so that the algorithm does not reconsider that possibility.

Node placement problems have been long investigated in the optimization field due to numerous applications in facility location, logistics, services, etc. Such problems are attracting again the attention of researchers now from the networking domain, and more especially from Wireless Mesh Networks (WMNs) field. Indeed, the placement of mesh routers nodes appears to be crucial for the performance and operability of WMNs, in terms of network connectivity and stability. However, node placement problems are known for their hardness in solving them to optimality, and therefore heuristics methods are approached to near-optimally solve such problems. In this work we evaluate the performance of different heuristic methods in order to judge on their suitability of solving mesh router nodes problem. We have selected methods from two different families, namely, local search methods (Hill Climbing and Simulated Annealing) and population-based methods (Genetic Algorithms). The former are known for their capability to exploit the solution space by constructing a path of visited solutions, while the later methods use a population of individuals aiming to largely explore the solution space. In both cases, a bi-objective optimization consisting in the maximization of the size of the giant component in the mesh routers network (for measuring network connectivity) and that of user coverage are considered. In the experimental evaluation, we have used a benchmark of instances--varying from small to large size--generated using different distributions of mesh node clients (Uniform, Normal, Exponential and Weibull).

Improvement on the computational efficiency of inverse queueing network analysis

Parameter identification is the process of constructing the mathematical model of a dynamic system based on measured dynamic responses. A non-classical search method namely genetic algorithm (GA) is employed in this study, which has several advantages over classical system identification techniques. Nevertheless, direct application of GA does not necessarily work well, particularly with regards to computational efficiency in fine-tuning when the solution approaches the optimal value. Thus, a local search (LS) method is introduced into the GA approach to improve this capability. Numerical example of a cantilever beam is presented to show the efficiency of the combined GA and LS method (GA-LS). The FE model of the cantilever beam is established with 10 beam elements. In this example, only the element properties associated with flapping motion are identified. The GA-LS method gives much better results than the GA method does. The mean error is reduced to 5.3% from 14.9% while to 14.5% from 47.2% for the maximum individual error. It is concluded that the GA-LS method is a more efficient search method than the GA method. The GA-LS method is then applied to identify the mass and stiffness parameters of a model helicopter blade using measured excitations and accelerations in this paper. Since the mass and stiffness of the model helicopter blade are not available, the only way to verify the identified parameters is to compare the ffequencies and mode shapes calculated from the identified parameters and the measured frequencies and modal shapes of the actual blade. The mean eiror of the frequencies is 0.9% and the maximum individual error 2.4%. The sequence of identified mode shapes is the same with measured mode shapes. Transactions on Engineering Sciences vol 43, © 2003 WIT Press, www.witpress.com, ISSN 1743-3533

Constraint Programming typically uses the technique of depth-first branch and bound as the method of solving optimization problems. Although this method can give the optimal solution, for large problems, the time needed to find the optimal can be prohibitive. This paper introduces a method for using local search techniques within a Constraint Programming framework, and applies this technique to vehicle routing problems. We introduce a Constraint Programming model for vehicle routing, and a system for integrating Constraint Programming and local search techniques. We then describe how the method can be accelerated by handling core constraints using fast local checks, while other more complex constraints are left to the constraint propagation system. We have coupled our local search method with a meta-heuristic to avoid the search being trapped in local minima. Several meta-heuristics are investigated ranging from a simple Tabu Search method to Guided Local Search. An empirical study over benchmark problems shows the relative merits of these techniques. Investigations indicate that the specific long-term memory technique used by Guided Local Search can be used as a diversification method for Tabu Search, resulting in significant benefit. Several new best solutions on the Solomon problems are found in relatively few iterations of our algorithm.

The idea of this extended foreword to the special issue on applied local search is threefold. Firstly, we provide a brief and fundamental description of what is nowadays called local search. Components which have meanwhile become an integral part of the classical aspects on simulated annealing, tabu search and genetic algorithms are reviewed. Furthermore, today tabu search can be considered as the major pillar of local search. Hence, attention is drawn to a couple of tabu search issues more recently developed such as the reverse elimination method and ejection chains. Secondly, local search based knowledge engineering is developed to constitute a substantial part of knowledge based information systems. Within this general setting local search will be considered as one particular paradigm of machine learning. Thirdly, we are going to introduce what is considered to be the main subject of this issue, local search applications. We briefly embed the contents of the subsequent papers and group them with respect to their particular methods and models within the above mentioned framework.

Today, there is a tendency towards promoting locally produced perishable products along the supply chain. Due to the short shelf-life, it is necessary to optimize collecting perishable goods to reduce wastage. The research attempt to solve a problem where internal and hired trucks are used to collect perishable goods. The proposed Vehicle Routing Problem (VRP) is different from the classical VRP and comprises both close and open routes. The problem is known as Close-Open Mixed Vehicle Routing Problem (COMVRP). The proposed model uses both internal and hired fleets with heterogeneous capacities. The internal trucks start from the multiple collecting centres and return to the origin point after completing the task. Further, hired trucks start from the third-party locations and return to the assigned collecting centre to store the goods. A Greedy algorithm-based heuristic was used to obtain the initial solutions. Moreover, several local search methods, including Guided Local Search (GLS), Simulated Annealing (SA), and Tabu Search (TS), were used to obtain the near-optimal solutions for the proposed problem. The research employed data from a real-world case study to conduct the numerical experiments in OR Tools. The results of the numerical experiments indicate that GLS outperforms in solving the proposed problem. Moreover, GLS obtained near-optimal solutions in reasonable computation time. Analysis showed that the proposed model is achieved more than 80% of truck capacity utilization. The proposed model assists industry practitioners to mitigate transport inefficiencies. Thereby it helps to minimize post-harvest wastage during the collection process. A further study compared real-driving distance and the Euclidean distance as input data. It is noted that when Euclidean distances are used, the algorithms fail to reach the optimal solutions.

Optimizing the fuel cycle cost through the optimization of nuclear reactor core loading patterns (LPs) involves multiple objectives and constraints, leading to a vast number of candidate solutions that cannot be explicitly solved. To advance the state of the art in core reload patterns, we have developed methods based on deep Reinforcement Learning (RL) for both single- and multi-objective optimization. Our previous research laid the groundwork for these approaches and demonstrated their ability to discover high-quality patterns within a reasonable time frame. On the other hand, Stochastic Optimization (SO) approaches are commonly used in the literature, but there is no rigorous explanation that shows which approach is better in which scenario. In this paper, we demonstrate the advantage of our RL-based approach, specifically using Proximal Policy Optimization (PPO) against the most commonly used SO-based methods: Genetic Algorithm, Parallel Simulated Annealing with mixing of states, and Tabu Search, as well as an ensemble-based method, i.e. the Prioritized replay Evolutionary and Swarm Algorithm. We found that the LP scenarios derived in this paper are amenable to a global search to identify promising research directions rapidly but then need to transition into a local search to exploit these directions efficiently and prevent getting stuck in local optima. PPO adapts its search capability via a policy with learnable weights, allowing it to function as both a global search method and a local search method. Subsequently, we compared all algorithms against PPO in long runs, which exacerbated the differences seen in the shorter cases. Overall, the work demonstrates the statistical superiority of PPO compared to the other considered algorithms.

MULTI-START LOCAL SEARCH FOR ONLINE SCHEDULING IN PARALLEL OPERATING THEATRE

Performance of simulated annealing-based heuristic for the unconstrained binary quadratic programming problem

Supervised Neural Network (NN) learning is a process in which input patterns and known targets are presented to a NN while it learns to recognize (classify, map, fit, etc.) them as desired. The learning is mathematically defined as an optimisation problem, i.e., an error function representing the differences between the desired and actual output, is being minimized (Bishop, 1995; Haykin, 1999). Because the most popular supervised learning techniques are gradient based (Backpropagation BP), they suffer from the so-called Local Minima Problem (Bishop, 1995). This has motivated the employment of Global Optimisation (GO) methods for the supervised NN learning. Stochastic and heuristic GO approaches including Evolutionary Algorithms (EA) demonstrated promising performance over the last decades (Smagt, 1994; Sexton et al., 1998; Jordanov & Georgieva, 2007; etc.). EA appeared more powerful than BP and its modifications (Sexton et al., 1998; Alba & Chicone 2004), but hybrid methods that combine the advantages of one or more GO techniques and local searches were proven to be even better (Yao, 1999; Rocha et al., 2003; Alba & Chicano, 2004; Ludemir et al., 2006). Hybrid methods were promoted over local searches and simple population based techniques in Alba & Chicone (2004). The authors compared five methods: two BP implementations (gradient descent and Levenberg-Marquardt), Genetic Algorithms (GA), and two hybrid methods, combining GA with different local methods. The methods were used for NN learning applied to problems arising in medicine. Ludemir et al. (2006) optimized simultaneously NN weights and topology with a hybrid method combining Simulated Annealing (SA), Tabu Search (TS) and BP. A set of new solutions was generated on each iteration by TS rules, but the best solution was only accepted according to the probability distribution as in conventional SA. Meanwhile, the topology of the NN was also optimized and the best solution was kept. Finally, BP was used to train the best NN topology found in the previous stages. The new methodology compared favorably with SA and TS on four classification and one prediction problems. Plaginakos et al. (2001) performed several experiments to evaluate various training methods – six Differential Evolution (DE) implementations (with different mutation operators), BP, BPD (BP with deflection), SA, hybridization of BP and SA (BPSA), and GA. They reported O pe n A cc es s D at ab as e w w w .in te ch w eb .o rg

Local search (LS) methods start from a point and use the gradient or objective function value to guide the search. Such methods are good in searching the neighborhood of a given solution (i.e., they are good at exploitation), but they are poor in exploration. Evolutionary Algorithms (EAs) are nature inspired populationbased search optimizers. They are good in exploration, but not as good at exploitation as LS methods. Thus, it makes sense to hybridize EAs with LS techniques to arrive at a method which benefits from both and, as a result, have good search ability. Broydon-Fletcher-Goldfarb-Shanno (BFGS) method is a gradient-based LS method designed for nonlinear optimization. It is an efficient, but expensive method. Adaptive Differential Evolution with Optional External Archive (JADE) is an efficient EA. Nonetheless, its performance decreases with the increase in problem dimension. In this paper, we present a new hybrid algorithm of JADE and BFGS, called Hybrid of Adaptive Differential Evolution and BFGS, or DEELS, to solve the unconstrained continuous optimization problems. The performance of DEELS is compared, in terms of the statistics of the function error values with JADE.

Many stochastic local search (SLS) methods rely on the manipulation of single solutions at each of the search steps. Examples are iterative improvement, iterated local search, simulated annealing, variable neighborhood search, and iterated greedy. These SLS methods are the basis of many state-of-the-art algorithms for hard combinatorial optimization problems. Often, several of these SLS methods are combined with each other to improve performance. We propose here a practical, unified structure that encompasses several such SLS methods. The proposed structure is unified because it integrates these metaheuristics into a single structure from which we can not only instantiate each of them, but we also can generate complex combinations and variants. Moreover, the structure is practical since we propose a method to instantiate actual algorithms for practical problems in a semi-automatic fashion. The method presented in this work implements a general local search structure as a grammar; an instantiation of such a grammar is a program that can be compiled into executable form. We propose to find the appropriate grammar instantiation for a particular problem by means of automatic configuration. The result is a semi-automatic system that, with little human effort, is able to generate powerful hybrid SLS algorithms.KeywordsStochastic local searchgeneralized local search structuregrammarautomatic algorithm design

Constraint programming and local search are two well known optimization technologies. In recent years, methods for combining these two technologies have been put forward, one of which advocates the use of constraint programming for searching the local neighborhood of the current solution. We present a search technique which improves on the performance of this constraint programming-based local search method and perform experiments on a variety of both simple and more complex combinatorial problems. We also demonstrate the benefit of combining local and complete search methods.