摘要 :
In most countries and for most livestock species, genomic evaluations are obtained from within-breed analyses. To achieve reliable breeding values, however, a sufficient reference sample size is essential. To increase this size, t...
展开
In most countries and for most livestock species, genomic evaluations are obtained from within-breed analyses. To achieve reliable breeding values, however, a sufficient reference sample size is essential. To increase this size, the use of multibreed reference populations for small populations is considered a suitable option in other species. Over decades, the separate breeding work of different pig breeding organizations in Germany has led to stratified subpopulations in the breed German Large White. Due to this fact and the limited number of Large White animals available in each organization, there was a pressing need for ascertaining if multi-subpopulation genomic prediction is superior compared with within-subpopulation prediction in pigs. Direct genomic breeding values were estimated with genomic BLUP for the trait "number of piglets born alive" using genotype data (Illumina Porcine 60K SNP BeadChip) from 2,053 German Large White animals from five different commercial pig breeding companies. To assess the prediction accuracy of within- and multi-subpopulation reference sets, a random 5-fold cross-validation with 20 replications was performed. The five subpopulations considered were only slightly differentiated from each other. However, the prediction accuracy of the multi-subpopulations approach was not better than that of the within-subpopulation evaluation, for which the predictive ability was already high. Reference sets composed of closely related multi-subpopulation sets performed better than sets of distantly related subpopulations but not better than the within-subpopulation approach. Despite the low differentiation of the five subpopulations, the genetic connectedness between these different subpopulations seems to be too small to improve the prediction accuracy by applying multi-subpopulation reference sets. Consequently, resources should be used for enlarging the reference population within subpopulation, for example, by adding genotyped females.
收起
摘要 :
In this paper, Subpopulation Firefly Algorithm is proposed for optimization of machining parameters in multi-pass turning and multi-pass face milling operations. Basic Firefly Algorithm is modified with the aim to avoid space of l...
展开
In this paper, Subpopulation Firefly Algorithm is proposed for optimization of machining parameters in multi-pass turning and multi-pass face milling operations. Basic Firefly Algorithm is modified with the aim to avoid space of local minimum and to meet the operation constraints in each iteration step. For that purpose, the following modifications are made: one firefly population is divided into two, a crossover operator is introduced and the searching for new design variables is continued until constraint functions are fulfilled. For turning operation, optimization is carried out for one objective: minimization of production cost. For face milling operation, multi-objective optimization is used for minimizing production cost and machining time, and maximizing profit rate at the same time. In both cases of multi-pass machining operations, optimization process implies meeting all operation constraints. For multi-pass turning operation, the best results from literature are confirmed with good convergence and low value of standard deviation. For multi-pass milling operation, better results are achieved compared with existing results from literature. The proposed algorithm showed capability of achieving global optimum for complex optimization problems.
收起
摘要 :
For evolutionary algorithms, the search data during evolution has attracted considerable attention and many kinds of data mining methods have been proposed to derive useful information behind these data so as to guide the evolutio...
展开
For evolutionary algorithms, the search data during evolution has attracted considerable attention and many kinds of data mining methods have been proposed to derive useful information behind these data so as to guide the evolution search. However, these methods mainly centered on the single objective optimization problems. In this paper, an adaptive differential evolution algorithm based on analysis of search data is developed for the multi-objective optimization problems. In this algorithm, the useful information is firstly derived from the search data during the evolution process by clustering and statistical methods, and then the derived information is used to guide the generation of new population and the local search. In addition, the proposed differential evolution algorithm adopts multiple subpopulations, each of which evolves according to the assigned crossover operator borrowed from genetic algorithms to generate perturbed vectors. During the evolution process, the size of each subpopulation is adaptively adjusted based on the information derived from its search results. The local search consists of two phases that focus on exploration and exploitation, respectively. Computational results on benchmark multi-objective problems show that the improvements of the strategies are positive and that the proposed differential evolution algorithm is competitive or superior to some previous multi-objective evolutionary algorithms in the literature. (C) 2016 Elsevier Inc. All rights reserved.
收起
摘要 :
Evolutionary algorithms have shown prominent performance in solving various kinds of multi-objective optimization problems (MOPs), but most of them only use single operator that is often sensitive to the characteristics of problem...
展开
Evolutionary algorithms have shown prominent performance in solving various kinds of multi-objective optimization problems (MOPs), but most of them only use single operator that is often sensitive to the characteristics of problems. As different operators have different search patterns, a proper combination of multiple operators can be more efficient and robust than using one single operator in solving complex problems. However, how to ensemble multiple operators based on their performances in optimization process is a challenging task. In the machine learning field, it is well known that AdaBoost can effectively ensemble multiple classifiers by giving the proper weights based on their classification errors. Inspired by this ensemble way, we propose a multi-operator ensemble (MOE) strategy based on multiple subpopulations for evolutionary multi-objective optimization. In the proposed strategy, the survival rate of each subpopulation after environmental selection is used to evaluate the performance of the operator, and then a credit assignment scheme is developed by using the weight update method in AdaBoost. Based on these credits, an emigration-immigration mechanism is designed to update the subpopulation that can adaptively reward or punish the computational resources for operators. Experimental results on three complex test suites demonstrate that the proposed MOE can significantly improve the performance of multi-objective evolutionary algorithms on different types of MOPs. (C) 2019 Elsevier B.V. All rights reserved.
收起
摘要 :
Multi-functional textile has been increasingly employed for various usages in sports, outdoor, city, casual and industrial materials. Due to shortening product life cycles of consumer era, textile batch dyeing scheduling problem t...
展开
Multi-functional textile has been increasingly employed for various usages in sports, outdoor, city, casual and industrial materials. Due to shortening product life cycles of consumer era, textile batch dyeing scheduling problem that can be modeled as the parallel batch processing machines with arbitrary job size, incompatible job family, different due date, and sequence-dependent setup time has increasingly complicated product mix, while smart production is needed. To migrate for Industry 4.0, this study aims to develop a multi-subpopulation genetic algorithm with heuristics embedded (MSGA-H) to minimize the makespan to improve the textile batch dyeing scheduling that is the bottleneck. In addition, an approach that combines the state-of-art methods of batch processing scheduling is developed for reference solutions. To estimate the validity of the proposed MSGA-H, an empirical study was conducted in a world leading vertically integrated textile manufacturer in Taiwan with different scenarios based on real settings. The results have shown practical viability of the proposed MSGA-H. This study concludes with a discussion of contributions and future research directions for smart production in emerging countries.
收起
摘要 :
Machine learning approaches have been effective in predicting adverse outcomes in different clinical settings. These models are often developed and evaluated on datasets with heterogeneous patient populations. However, good predic...
展开
Machine learning approaches have been effective in predicting adverse outcomes in different clinical settings. These models are often developed and evaluated on datasets with heterogeneous patient populations. However, good predictive performance on the aggregate population does not imply good performance for specific groups. In this work, we present a two-step framework to 1) learn relevant patient subgroups, and 2) predict an outcome for separate patient populations in a multi-task framework, where each population is a separate task. We demonstrate how to discover relevant groups in an unsupervised way with a sequence-to-sequence autoencoder. We show that using these groups in a multi-task framework leads to better predictive performance of in-hospital mortality both across groups and overall. We also highlight the need for more granular evaluation of performance when dealing with heterogeneous populations.
收起
摘要 :
Tools such as protein immunoblotting have proven benefits for investigating T lymphocyte signaling but have several drawbacks such as the number of cells required and the difficulty of distinguishing subset-specific differences wi...
展开
Tools such as protein immunoblotting have proven benefits for investigating T lymphocyte signaling but have several drawbacks such as the number of cells required and the difficulty of distinguishing subset-specific differences without expensive and invasive cell sorting. Recent advances in immunology and the identification of T lymphocyte sub-populations making up only a very small fraction of the total population highlight the importance of studying signaling in those small subsets in a feasible, cost-effective, high-throughput manner. To this end, we have developed a simplified protocol to study both intracellular phosphorylation patterns of important signal transduction molecules concomitantly with T cell surface marker expression. A multi-parametric analysis may allow the quantification of the phosphorylation of up to five signaling molecules in CD4 and CD8 T lymphocytes and their na?ve, central memory, effector memory, and TEMRA subsets. This enables precise identification of subset-specific signaling and alterations of signaling pathways in physiological and pathological situations. The importance of such detailed analysis is discussed.
收起
摘要 :
During the development of new therapies, it is not uncommon to test whether a new treatment works better than the existing treatment for all patients who suffer from a condition (full population) or for a subset of the full popula...
展开
During the development of new therapies, it is not uncommon to test whether a new treatment works better than the existing treatment for all patients who suffer from a condition (full population) or for a subset of the full population (subpopulation). One approach that may be used for this objective is to have two separate trials, where in the first trial, data are collected to determine if the new treatment benefits the full population or the subpopulation. The second trial is a confirmatory trial to test the new treatment in the population selected in the first trial. In this paper, we consider the more efficient two-stage adaptive seamless designs (ASDs), where in stage 1, data are collected to select the population to test in stage 2. In stage 2, additional data are collected to perform confirmatory analysis for the selected population. Unlike the approach that uses two separate trials, for ASDs, stage 1 data are also used in the confirmatory analysis. Although ASDs are efficient, using stage 1 data both for selection and confirmatory analysis introduces selection bias and consequently statistical challenges in making inference. We will focus on point estimation for such trials. In this paper, we describe the extent of bias for estimators that ignore multiple hypotheses and selecting the population that is most likely to give positive trial results based on observed stage 1 data. We then derive conditionally unbiased estimators and examine their mean squared errors for different scenarios.(c) 2015 The Authors. Statistics in Medicine Published by JohnWiley & Sons Ltd.
收起
摘要 :
Higher accuracy in cluster failure prediction can ensure the long-term stable operation of cluster systems and effectively alleviate energy losses caused by system failures. Previous works have mostly employed BP neural networks (...
展开
Higher accuracy in cluster failure prediction can ensure the long-term stable operation of cluster systems and effectively alleviate energy losses caused by system failures. Previous works have mostly employed BP neural networks (BPNNs) to predict system faults, but this approach suffers from reduced prediction accuracy due to the inappropriate initialization of weights and thresholds. To address these issues, this paper proposes an improved arithmetic optimization algorithm (AOA) to optimize the initial weights and thresholds in BPNNs. Specifically, we first introduced an improved AOA via multi-subpopulation and comprehensive learning strategies, called MCLAOA. This approach employed multi-subpopulations to effectively alleviate the poor global exploration performance caused by a single elite, and the comprehensive learning strategy enhanced the exploitation performance via information exchange among individuals. More importantly, a nonlinear strategy with a tangent function was designed to ensure a smooth balance and transition between exploration and exploitation. Secondly, the proposed MCLAOA was utilized to optimize the initial weights and thresholds of BPNNs in cluster fault prediction, which could enhance the accuracy of fault prediction models. Finally, the experimental results for 23 benchmark functions, CEC2020 benchmark problems, and two engineering examples demonstrated that the proposed MCLAOA outperformed other swarm intelligence algorithms. For the 23 benchmark functions, it improved the optimal solutions in 16 functions compared to the basic AOA. The proposed fault prediction model achieved comparable performance to other swarm-intelligence-based BPNN models. Compared to basic BPNNs and AOA-BPNNs, the MCLAOA-BPNN showed improvements of 2.0538 and 0.8762 in terms of mean absolute percentage error, respectively.
收起
摘要 :
In this paper, we propose a new economic dispatch model with random wind power, demand response and carbon tax. The specific feature of the demand response model is that the consumer's electricity demand is divided into two parts:...
展开
In this paper, we propose a new economic dispatch model with random wind power, demand response and carbon tax. The specific feature of the demand response model is that the consumer's electricity demand is divided into two parts: necessary part and non-essential part. The part of the consumer's participation in the demand response is the non-essential part of the electricity consumption. The optimal dispatch objective is to obtain the minimum total cost (fuel cost, random wind power cost and emission cost) and the maximum consumer's non-essential demand response benefit while satisfying some given constraints. In order to solve the optimal dispatch objective, a multi-subpopulation bat optimization algorithm (MSPBA) is proposed by using different search strategies. Finally, a case of an economic dispatch model is given to verify the feasibility and effectiveness of the established mathematical model and proposed algorithm. The economic dispatch model includes three thermal generators, two wind turbines and two consumers. The simulation results show that the proposed model can reduce the consumer's electricity demand, reduce fuel cost and reduce the impact on the environment while considering random wind energy, non-essential demand response and carbon tax. In addition, the superiority of the proposed algorithm is verified by comparing with the optimization results of CPLEX+YALMIP toolbox for MATLAB, BA, DBA and ILSSIWBA.
收起