Parameter Setting Approach Research Articles

Performance improvement strategies on Cuckoo Search algorithms for solving the university course timetabling problem

The university course timetabling problem (UCTP) arises every academic year and must be solved by academic staff with/without a course timetabling tool. A Hybrid Self-adaptive Cuckoo Search-based Timetabling (HSCST) tool has been developed for minimising the total university operating costs. The HSCST tool was applied to solve eleven problem instances obtained from the Faculty of Engineering, Naresuan University. The performance improvements of the Cuckoo Search (CS) algorithm embedded within the proposed tool were demonstrated using three strategies: parameter setting approaches (static and adaptive), movement strategies (Lévy flights and Gaussian random walks), and local search hybridisation techniques. Sequential computational experiments were designed and conducted to investigate the efficiency of the three proposed strategies. The statistical analysis on the computational results suggested that the proposed algorithms significantly outperformed the conventional CS, Particle Swarm Optimisation (PSO), and hybrid PSO for all problem instances.

A survey of dynamic parameter setting methods for nature-inspired swarm intelligence algorithms

Parameter settings for nature-inspired optimization algorithms are essential for their effective performance. Evolutionary algorithms and swarm intelligence algorithms are prominent types of nature-inspired optimization. There are comprehensive reviews of parameter setting techniques for evolutionary algorithms. Counterparts providing an overview of parameter setting techniques for swarm intelligence algorithms are needed also. Therefore, in this paper, we provide a critical and comprehensive review, focusing in particular on dynamic parameter setting techniques. The paper describes a variety of swarm intelligence algorithms and parameter setting approaches that have been applied to them. This review simplifies the selection of parameter setting techniques for each algorithm by collecting them in a single document and classifying them under a taxonomy. Recommendations for parameter setting approach selection are provided in this review. We explore the open problems related to dynamic parameter setting techniques for swarm intelligence optimization and discuss the trade-off between run-time computation and flexibility of these algorithms.

Effect of parameter setting and spectral normalization approach on study of matrix effect by laser induced breakdown spectroscopy of Ag–Zn binary composites

The complex nature of laser-material interaction causes non-stoichiometric ablation of alloy samples. This is attributed to matrix effect, which reduces analyzing capability. To address this issue, the analytical performance of three different normalization methods, namely normalization with background, internal normalization and three point smoothing techniques at different parameter settings is studied for quantification of Ag and Zn by Laser induced breakdown spectroscopy (LIBS). The LIBS spectra of five known concentration of silver zinc binary composites have been investigated at various laser irradiances (LIs). Calibration curves for both Ag(I) line (4d105s2S1/2 → 4d105p2P1/2 at 338.28 nm) and Zn(I) line (4s5s3S1 → 4s4p3P2 at 481.053 nm) have been determined at LI of 5.86 × 1010 W cm−2. Slopes of these calibration curves provide the valuation of matrix effect in the Ag–Zn composites. With careful sample preparation and normalization after smoothing at optimum parameter setting (OPS), the minimization of sample matrix effect has been successfully achieved. A good linearity has been obtained in Ag and Zn calibration curve at OPS when normalized the whole area of spectrum after smoothing and the obtained coefficients of determination values were R2 = 0.995 and 0.998 closer to 1. The results of matrix effect have been further verified by analysis of plasma parameters. Both plasma parameters showed no change with varying concentration at OPS. However, at high concentration of Ag, the observed significant changes in both plasma parameters at common parameter setting PS-1 and PS-2 were the gesture of matrix effect. In our case, the better analytical results were obtained at smoothing function with optimized parameter setting that indicates it is more efficient than normalization with background and internal normalization method.

An Experimental Study in Adaptive Kernel Selection for Bayesian Optimization

Bayesian Optimization has been widely used along with Gaussian Processes for solving expensive-to-evaluate black-box optimization problems. Overall, this approach has shown good results, and particularly for parameter tuning of machine learning algorithms. Nonetheless, Bayesian Optimization has to be also configured to achieve the best possible performance, being the selection of the kernel function a crucial choice. This paper investigates the convenience of adaptively changing the kernel function during the optimization process, instead of fixing it a priori. Six adaptive kernel selection strategies are introduced and tested in well-known synthetic and real-world optimization problems. In order to provide a more complete evaluation of the proposed kernel selection variants, two major kernel parameter setting approaches have been tested. According to our results, apart from having the advantage of removing the selection of the kernel out of the equation, adaptive kernel selection criteria show a better performance than fixed-kernel approaches.

Sorption parameter setting approaches for radioactive waste disposal considering perturbation effects: sorption reduction factors for organics

Sorption parameter setting approaches for radioactive waste disposal considering perturbation effects: sorption reduction factors for organics

Melia: A MapReduce Framework on OpenCL-Based FPGAs

MapReduce, originally developed by Google for search applications, has recently become a popular programming framework for parallel and distributed environments. This paper presents an energy-efficient architecture design for MapReduce on Field Programmable Gate Arrays (FPGAs). The major goal is to enable users to program FPGAs with simple MapReduce interfaces, and meanwhile to embrace automatic performance optimizations within the MapReduce framework. Compared to other processors like CPUs and GPUs, FPGAs are (re-)programmable hardware and have very low energy consumption. However, the design and implementation of MapReduce on FPGAs can be challenging: firstly, FPGAs are usually programmed with hardware description languages, which hurts the programmability of the MapReduce design to its users; secondly, since MapReduce has irregular access patterns (especially in the reduce phase) and needs to support user-defined functions, careful designs and optimizations are required for efficiency. In this paper, we design, implement and evaluate Melia , a MapReduce framework on FPGAs. Melia takes advantage of the recent OpenCL programming framework developed for Altera FPGAs, and abstracts FPGAs behind the simple and familiar MapReduce interfaces in C. We further develop a series of FPGA-centric optimization techniques to improve the efficiency of Melia, and a cost- and resource-based approach to automate the parameter settings for those optimizations. We evaluate Melia on a recent Altera Stratix V GX FPGA with a number of commonly used MapReduce benchmarks. Our results demonstrate that 1) the efficiency and effectiveness of our optimizations and automated parameter setting approach, 2) Melia can achieve promising energy efficiency in comparison with its counterparts on CPUs/GPUs on both single-FPGA and cluster settings.

A fast nonlocally centralized sparse representation algorithm for image denoising

The sparsity from self-similarity properties of natural images, which has received significant attention in the image processing community of researchers, is widely applied for image denoising. The recently proposed nonlocally centralized sparse representation (NCSR) algorithm that takes advantage of the sparse representations (SRs) and the nonlocal estimate of sparse coefficients (NESCs) has shown promising results with respect to noise reduction. Despite successful combination of the above two techniques, the iterative dictionary learning and the nonlocal estimate of unknown sparse coefficients make this algorithm computationally demanding, which largely limits its applicability in many applications. To address this problem, a fast version of the NCSR algorithm called FNCSR algorithm, which is based on pre-learned dictionary and adaptive parameter setting approaches, was proposed in this paper. Specifically, we adopted the same dictionary learning approach, i.e, the K-means and principal component analysis (PCA), with the NCSR algorithm to obtain a dictionary for each image in a selected image dataset including high-quality natural and texture images. Then we applied PNSR index to objectively assess the image quality of the reconstructed images using these dictionaries throughout the image dataset. The dictionary providing the best average reconstructed quality was selected as fixed dictionary, i.e., the pre-learned dictionary, for sparse coding throughout the iterative denoising process, which implies that it no longer requires dictionary learning procedure within the framework of the proposed FNCSR algorithm, resulting in greatly decreased execution time. In order to further improve computational efficiency, we employed quality-aware features and support vector regression (SVR) technique to build a fast noise level estimator (NLE) to estimate the noise level from a single noisy image. The parameters related to the NESC, i.e., the search window and the search step, which influences the computational performance of the NCSR algorithm strongly, were chosen automatically according to the estimated noise level. Compared to the original NCSR algorithm, these modifications lead to substantial benefits in computational efficiency (a performance gain of about 90% can be achieved) without sacrificing image quality too much (the largest decline is less than 0.55dB and 0.014 in terms of PSNR and SSIM indices). Compared with other state-of-the-art denoising algorithms, experimental results show that the proposed FNCNR algorithm also achieves comparable performance in terms of both quantitative measures and visual quality.

Bayesian network as an adaptive parameter setting approach for genetic algorithms

Parameter setting currently ranks among the most actively researched topics in the evolutionary algorithm (EA) community. This can be explained by the major impact EA parameters have on search performance. However, parameter setting has been shown to be both problem dependent and evolution dependent. Moreover, because parameters interact in complex ways, developing an efficient and beneficial parameter setting approach is not an easy feat, and no broadly recognized solution has emerged to date. In this paper, we borrow the notion of parameter adaptation with the objective of addressing the parameter setting dependencies mentioned above, using a strategy based on a Bayesian network. The adaptive framework is elaborated for a steady-state genetic algorithm (SSGA) to control nine parameters. To judge parameter state productivities, we consider the population’s fitness improvement, as well as exploration/exploitation balance management. The performance of this proposal, a Bayesian network for genetic algorithm parameter adaptation (BNGA), is assessed based on the CEC’05 benchmark. BNGA is compared to a static parameter setting, a naive approach, three common adaptive systems (PM, AP, and FAUC–RMAB), and two state-of-the-art EAs (CMA-ES and G-CMA-ES). Our results statistically demonstrate that the performance of BNGA is equivalent to that of FAUC–RMAB, CMA-ES, and G-CMA-ES, and overall is superior to that of all the other SSGA parameter setting approaches. However, these results also reveal that all the approaches considered have great difficulty finding global optima in a multimodal problem set. This suggests a lack of complementarity and/or synergy among parameter states.

Effective Utilization of Router Buffer by Threshold Parameter Setting Approach in RED

Active Queue Management (AQM) provide solution to Network congestion of Internet. Random Early Detection(RED) is the first well known AQM recommended by Internet Engineering Task Force (IETF) used for congestion avoidance for last three decades. RED has some disadvantages like hard tuning of control parameters, inefficient congestion notification, insensitivity to variation of traffic load. In this research work above issues are addressed and a mixed approach of threshold parameter tuning in terms of router buffer space and inclusion of traffic load in congestion notification along with average size is used. The approach is given the name LTRED, here L is for length of buffer and T stands for Threshold. In this research work, impact of variation of router queue size in terms of bandwidth under different traffic load scenario is observed and compared with standard AQM’s like RED, ARED and AVQ. Extensive simulations using ns-2 simulator demonstrates that LTRED outperforms others in terms of effective utilization of router buffer space, less packet loss, high goodput and high link utilization.

Optimal Inertial Support from the Variable Speed Wind Turbines using Particle Swarm Optimization

Variable Speed Wind Turbines (VSWT) are usually devoid of inertial response, a natural phenomenon observed in Synchronous Generators (SG). With increased generation share from wind energy, most of which employ VSWTs, the major concern for the utility operators is the reduction in grid inertia. Recently, various modifications in the outer control loop of VSWTs are suggested, which usually establish a link between the grid frequency deviations and the generator torque/active power to mimic the inertial response observed in SGs. Since the inertial support is synthetic and is the result of additional control loop, the parameters of the auxiliary control loop has profound effect on the emulated inertial response. In this work, a simple parameter selection approach based on Particle Swarm Optimization (PSO) is proposed. The objective is to enable the optimum extraction of stored kinetic energy while limiting the effects of consequent under-production period. The proposed approach for parameter setting is compared to two different approaches and found to be more effective in both limiting the frequency nadir and post-support disturbances.

Wind power modeling using harmony search with a novel parameter setting approach

A significant portion of the world’s renewable energy is being provided by wind farms. The produced power of a wind farm in not always a constant value because it is highly dependent on the environmental conditions. As the amount of wind power rapidly increases, wind power estimation will be an important issue in electricity markets. A number of wind power estimation methodologies have been developed that are competing to provide the best estimation to the electric power industry. In this paper, a harmony search (HS) algorithm is proposed for modeling the power of a wind farm by finding the optimal weighting factors of a mathematical model. HS with a novel parameter setting approach, named HS-NPSA, is also developed to eliminate the tedious parameter assigning efforts of HS algorithm. In order to evaluate the usefulness of the proposed methodologies, the performance of HS-based algorithms is compared with those of the other heuristic techniques in terms of mean absolute percentage error (MAPE). Simulation results confirm the superiority of HS-NPSA algorithm for wind power modeling.

A PSO-Based Fuzzy Long-Term Multi-Objective Optimization Approach for Placement and Parameter Setting of UPFC

This paper presents a Particle Swarm Optimization (PSO)-based fuzzy multi-objective methodology for optimal locating and parameter setting of Unified Power Flow Controller (UPFC) in a power system for a long-term period. One of the profits obtained by UPFC placement in a transmission network is the reduction in total generation cost due to its ability to change the power flow pattern in the network. Considering this ability, UPFC can be also used to remove or at least mitigate the congestion in transmission networks. The other issue in a power system is voltage violation which could even render the optimal power flow problem infeasible to be solved. Voltage violation could be also mitigated by proper application of UPFC in a transmission system. These objectives are considered simultaneously in a unified objective function for the proposed optimization algorithm. At first, these objectives are fuzzified and designed to be comparable against each other and then they are integrated and introduced to a PSO Algorithm to find the solution which maximizes the value of integrated objective function in a 3-year planning horizon, considering the load growth. A power injection model is adopted for UPFC. Unlike the most previous works in this field the parameters of UPFC are set for each load level to avoid inconvenient rejection of more optimal solutions. IEEE Reliability Test System is used as an illustrative example to show the effectiveness of the proposed method.

Efficient and optimal approach for location and parameter setting of multiple unified power flow controllers for a deregulated power sector

This study presents an efficient approach to optimal allocation of unified power flow controller (UPFC) in a pool and bilateral power market. The optimisation technique has been used with an objective to maximising social welfare (SW) by suitable location, number and rating of the UPFC in power system working under a deregulated environment. The proposed method accurately evaluates SW by determining locational marginal pricing based on the marginal cost theory. The impact of the UPFC on system real power losses has also been investigated. The effectiveness of the proposed optimal technique has been tested on the 39-bus New England system with addition of four bilateral transactions.

Tuning a parametric Clarke–Wright heuristic via a genetic algorithm

Almost all heuristic optimization procedures require the presence of a well-tuned set of parameters. The tuning of these parameters is usually a critical issue and may entail intensive computational requirements. We propose a fast and effective approach composed of two distinct stages. In the first stage, a genetic algorithm is applied to a small subset of representative problems to determine a few robust parameter sets. In the second stage, these sets of parameters are the starting points for a fast local search procedure, able to more deeply investigate the space of parameter sets for each problem to be solved. This method is tested on a parametric version of the Clarke and Wright algorithm and the results are compared with an enumerative parameter-setting approach previously proposed in the literature. The results of our computational testing show that our new parameter-setting procedure produces results of the same quality as the enumerative approach, but requires much shorter computational time.

Commentary on “The Acquisition of Inflection: A Parameter Setting Approach”

Click to increase image sizeClick to decrease image size The Ohio State University Department of Linguistics has given us permission to publish the overlapping portions of the article.

The Acquisition of Inflection: A Parameter-Setting Approach

First written in 1986, prior to the many findings concerning the optionality of finiteness and the root infinitive phenomenon, this article attempts to extend the parameter-setting model of grammatical development to the acquisition of inflectional morphology. I propose that the Stem Parameter, which states that a stem is/is not a well-formed word in the language, is set early, and that a positive vs. negative setting directly affects the timing and manner in which children acquire inflectional morphemes. Related to this, I propose that the distinction between core and peripheral grammar (Chomsky (1981)) provides a complexity metric for grammatical development.

A dynamic evolutionary mechanism for mixed-production

This paper describes a new dynamic evolutionary mechanism which assists process engineers in devising efficient processes for manufacturing high quality items where the mixed production approach is adopted. An adaptive system, including the use of genetic algorithms (GA) as a dynamic searching mechanism, is designed in order to maximize the stability of the quality control in the mixed production processes. GA is an effective approach in optimization as it is able to alter manufacturing variables so as to reach a global optimum in complex production processes such as multiple quality chains. The choice of the GA operators and its parameters, however, is a significant problem and inappropriate selection of chromosome structure can lead to poor performance. In order to deal with these issues, a dynamic parameter and operator setting approach with a mechanism based on quality control chart theory, is proposed. The approach allows a trade-off between exploration and exploitation processes in the search. The mechanism applies evolution evidence to supervise and adjust the GA parameter settings at run time. A prototype system has been implemented and applied to optimization problems in multiple quality chains. The experimental results have revealed that the dynamic setting approach can improve the performance of a GA process in multiple quality chains. The results also established that the dynamic setting approach is superior to a static one.

Parameter setting in a bio-inspired model for dynamic flexible job shop scheduling with sequence-dependent setups

This paper addresses the parameter modelling and optimisation issues in the application of a bio-inspired model for the scheduling of dynamic flexible job shop with sequence-dependent setups. The study sets the model parameters using two steps, termed mapping and tuning. Mapping establishes a set of coefficients to link the model parameters with the scheduling problem characteristics. Tuning determines good values of these coefficients and is then used to compute the model parameters. Such a tuning procedure is accomplished by extensive computational experiments and statistical analyses. A data set from semiconductor manufacturing was used to show the effectiveness of the parameter setting approach. The performance of the proposed multiagent model was compared with that of another scheduling method which is based on dispatching rules. It is concluded that the proposed parameter setting method is effective and worth considering when applying bio-inspired division of labour to dynamic manufacturing scheduling. [Received 5 October 2006; Revised 2 January 2007; Accepted 1 March 2007]

Fully automated intensity compensation for confocal microscopic images

One well-recognized problem in three-dimensional (3D) confocal microscopic images is that the intensities in deeper slices are generally weaker than those in shallower slices. The loss of intensity with depth hampers both qualitative observation and quantitative measurement of specimens. Two major types of methods exist to compensate for this intensity loss: the first is based on the geometrical optics inside the specimen, and the second applies an empirical parametric intensity decay function (IDF) of depth. A common feature shared by both methods is that they are parameter-dependent. However, for the optics-based method there are as yet no fully automated parameter-setting approaches; and for the IDF method the traditional profile-fitting approach cannot provide proper parameters if the presumed IDF model does not match the experimental intensity-depth profile of the 3D image. In this paper, we propose a novel maximum-entropy (ME) approach to fully automated parameter-setting. In principle the ME approach is suitable for any compensation method as long as it is parameter-dependent. The basic assumption is that without intensity loss an ideal 3D image should be generally homogeneous with respect to depth and this axial homogeneity can be represented by the entropy of a normalized intensity-depth profile. Experiments on real confocal images showed that such a profile was consistent with visual evaluation of axial intensity homogeneity and that the ME approach could provide proper parameters for both compensation methods mentioned above. Moreover, for the IDF method, experiments on both real and simulated data showed that the ME approach could provide more precise parameters than with traditional profile-fitting. The Appendix provides a proof that under certain conditions the global maximization of the profile-entropy is guaranteed.

PARAMETER SETTING IN LANGUAGE ACQUISITION

PARAMETER SETTING IN LANGUAGE ACQUISITION. Dalila Ayoun. New York: Continuum, 2003. Pp. x + 212. $110.00 cloth.This book presents an overview of the parameter setting theory of learnability in first (L1) and second (L2) language acquisition within the generative linguistic framework. It also attempts to challenge and refine common assumptions underlying the model. The book comprises five central chapters as well as short introductory and concluding chapters. The introductory chapter summarizes the general aim of the book and the specific aims of the chapters to follow. In chapter 2, Ayoun presents historical background on the concept of parameter throughout different versions of generative linguistics and distinguishes the standard notion of parameter from the notions of associated clusters of structures, microparameters (referring to structures), and macroparameters (which apply to a family of typologically different languages). This chapter also reviews the concept of parameter setting for language changes, creole formation, computational linguistics, and neurolinguistics, ending with a brief discussion of Universal Grammar and the Critical Period Hypothesis. In brief, this chapter argues that the parameter setting approach is a model worthy of further development and refinement, capable of explaining and predicting a wide range of phenomena in linguistic theory and its applications despite misunderstandings and lack of clarity in the field.