An artificial neural network that models human decision making

The Neural Logic Network (Neulonet) system models a wide range of human decision making behaviors by combining the strengths of rule based expert systems and neural networks. Neulonet differs from other neural networks by having an ordered pair of numbers associated with each node and connection, as shown. Let Q be the output node and P₁, P, …, P_N, be input nodes. Also, let values associated with the node P_i, be denoted by (a_i, b_i,), and the weight for the connection from P_i, to Q be (α_i,β_i,). Each node's ordered pair takes one of three values-(1,0) for true, (0,1) for false, or (0,0) for “don't know”; (1,1) is undefined     

An integrated decision making approach for ERP system selection

Enterprise resource planning (ERP) systems have gained major prominence by enabling companies to streamline their operations, leverage and integrate business data process. In order to implement an ERP project successfully, it is necessary to select an ERP system which can be aligned with the needs of the company. Thus, a robust decision making approach for ERP software selection requires both company needs and characteristics of the ERP system and their interactions to be taken into account. This paper develops a novel decision framework for ERP software selection based on quality function deployment (QFD), fuzzy linear regression and zero–one goal programming. The proposed framework enables both company demands and ERP system characteristics to be considered, and provides the means for incorporating not only the relationships between company demands and ERP system characteristics but also the interactions between ERP system characteristics through adopting the QFD principles. The presented methodology appears as a sound investment decision making tool for ERP systems as well as other information systems. The potential use of the proposed decision framework is illustrated through an application.     

选择性聚类融合新方法研究

针对传统选择性聚类融合算法不能消除劣质聚类成员的干扰以及聚类准确性不高等问题,提出了一种新的选择性加权聚类融合算法。算法中提出了基于聚类有效性评价方法的参照成员选择方法和联合聚类质量以及差异度的选择策略,然后还提出了基于容错关系信息熵的属性重要性加权方法。新算法有效地克服了传统选择性聚类融合算法的缺点,消除了劣质聚类成员的干扰,提高了聚类的准确性。大量的对比实验结果表明了算法的有效,且性能显著提高。     

An AHP/DEA methodology for ranking decision making units

This paper presents a two-stage model for fully ranking organizational units where each unit has multiple inputs and outputs. In the first stage, the Data Envelopment Analysis (DEA) is run for each pair of units separately. In the second stage, the pairwise evaluation matrix generated in the first stage is utilized to rank scale the units via the Analytical Hierarchical Process (AHP). The consistency of this AHP/DEA evaluation can be tested statistically. Its goodness of fit with the DEA classification (to efficient/inefficient) can also be tested using non-parametric tests. Both DEA and AHP are commonly used in practice. Both have limitations. The hybrid model AHP/DEA takes the best of both models, by avoiding the pitfalls of each. The nonaxiomatic utility theory limitations of AHP are irrelevant here: since we are working with given inputs and outputs of units, no subjective assessment of a decision maker evaluation is involved. AHP/DEA ranking does not replace the DEA classification model, rather it furthers the analysis by providing full ranking in the DEA context for all units, efficient and inefficient.     

A data clustering algorithm for stratified data partitioning in artificial neural network

The statistical properties of training, validation and test data play an important role in assuring optimal performance in artificial neural networks (ANNs). Researchers have proposed optimized data partitioning (ODP) and stratified data partitioning (SDP) methods to partition of input data into training, validation and test datasets. ODP methods based on genetic algorithm (GA) are computationally expensive as the random search space can be in the power of twenty or more for an average sized dataset. For SDP methods, clustering algorithms such as self organizing map (SOM) and fuzzy clustering (FC) are used to form strata. It is assumed that data points in any individual stratum are in close statistical agreement. Reported clustering algorithms are designed to form natural clusters. In the case of large multivariate datasets, some of these natural clusters can be big enough such that the furthest data vectors are statistically far away from the mean. Further, these algorithms are computationally expensive as well. We propose a custom design clustering algorithm (CDCA) to overcome these shortcomings. Comparisons are made using three benchmark case studies, one each from classification, function approximation and prediction domains. The proposed CDCA data partitioning method is evaluated in comparison with SOM, FC and GA based data partitioning methods. It is found that the CDCA data partitioning method not only perform well but also reduces the average CPU time.     

On the performance of the LP-guided Hopfield network-genetic algorithm

This paper proposes a linear programming (LP)-guided Hopfield-genetic algorithm for a class of combinatorial optimization problems which admit a 0–1 integer linear programming. The algorithm modifies the updating order of the binary Hopfield network in order to obtain better performance of the complete hybrid approach. We theoretically analyze several different updating orders proposed. We also include in the paper a novel proposal to guide the Hopfield network using the crossover and mutation operators of the genetic algorithm. Experimental evidences that show the good performance of the proposed approach in two different combinatorial optimization problems are also included in the paper.     

A multidisciplinary ensemble algorithm for clustering heterogeneous datasets

Neural Computing and Applications - Clustering is a commonly used method for exploring and analysing data where the primary objective is to categorise observations into similar clusters. In recent...     

一种改进的ART2网络学习算法

分析了现有ART2网络存在的问题,提出了一种改进的ART2算法。该算法首先利用样本数据自身来初始化权值,然后按照同一类中的数据点到其聚类中心的距离之和越小(即类内偏差越小),聚类效果越好的原则来设计特征表示场和类别表示场之间的权值修正公式,最后通过比较输入样本和聚类中心的模来有效地利用模式的幅度信息。分析证明了该算法不仅能有效解决模式漂移问题、充分利用幅度信息,而且能提高聚类速度。     

基于决策树算法确定一个最小神经网络

神经网络隐层神经元的个数对于网络的性能有着重要的影响,通常情况下,对于一个特定问题来说,没有一个确定的方法来决定隐含层到底应该有多少个神经元,一般采用试探的方法通过多次实验来达到理想效果.在分类问题中,决策树和神经网络的结构有着一定的关联性,通过把决策树映射到神经网络结构中来确定隐层神经元的个数的方法能够有效地设计神经网络的结构,从而提高训练的效率并达到良好的分类效果.实验结果表明,该方法能够得到一个有着良好识别率的最小神经网络.方法简单有效,直观且易于操作.     

An integrated algorithm for performance optimization of neurosurgical ICUs

This study presents an integrated simulation and data envelopment analysis (DEA) approach to increase the quality of service in a neurosurgical intensive care unit (ICU). The aim of this study is to capture the main factors which have negatively affects the patients’ satisfactions and figure out their optimized levels. In order to avoid any interruption in ICU's routine functions and being able to convince the hospital's principals about the project's outcomes, a simulation model is developed and run for different scenarios. Then DEA is used to compare the outputs of different scenarios. These scenarios are generated by observing the effects of various parameters such as lengthening or shortening treatment times, decreasing or increasing patient volumes and removing or adding staff members. As the best of our knowledge, this is the first study that presents an integrated approach based on computer simulation and DEA to concurrently incorporate the stated factors and parameters for optimization of complex ICUs in developing countries. Therefore, the results of this study are more precise and reliable than previous studies because of concurrent consideration of the stated factors.     

Intelligent decision making in multi-agent robot soccer system through compounded artificial neural networks

This paper proposes a two-stage approach using artificial neural networks for the intelligent decision-making by the robots in a MiroSot small league. The first stage involves the use of an evolutionary algorithm for getting a rough estimate of the neural network weight matrices. The proposed approach is then generalized to the case of quick, intelligent and accurate decision-making in the case of a robot soccer system with robots utilizing the concept of compounded artificial neural networks. In the proposed approach a soccer field is divided into three zones so that the decision of the robots depends on the zone of the ball at any instant. The concept of a forward robot is also introduced in this paper to enhance the accuracy of the decision-making with the global strategy of advancing towards the goal area of the opponent for scoring a goal. Simulation results indicate that the proposed techniques are very effective in taking intelligent decision-making in a multi-agent robot soccer system in MiroSot small league as well as middle league.     

An improved artificial bee colony algorithm based on whale optimization algorithm for data clustering

Multimedia Tools and Applications - Data clustering is one of the branches of unsupervised learning and it is a process whereby the samples are divided into categories whose members are similar to...     

An integrated fuzzy multi criteria group decision making approach for ERP system selection

This paper aims to ease group decision-making by using an integration of fuzzy AHP (analytic hierarchy process) and fuzzy TOPSIS (technique for order preference by similarity to ideal solution) and its application to software selection of an electronic firm. Firstly, priority values of criteria in software selection problem have been determined by using fuzzy extension of AHP method. Fuzzy extension of AHP is suggested in this paper because of little computation time and much simpler than other fuzzy AHP procedures. Then, the result of the fuzzy TOPSIS model can be employed to define the most appropriate alternative with regard to this firm's goals in uncertain environment. Fuzzy numbers are presented in all phases in order to overcome any vagueness in decision making process. The final decision depends on the degree of importance of each decision maker so that wrong degree of importance causes the mistaken result. The researchers generally determine the degrees of importance of each decision maker according to special characteristics of each decision maker as subjectivity. In order to overcome this subjectivity in this paper, the judgments of decision makers are degraded to unique decision by using an attribute based aggregation technique. There is no study about software selection using integrated fuzzy AHP-fuzzy TOPSIS approach with group decision-making based on an attribute based aggregation technique. The results of the proposed approach and the other approaches are compared. Results indicate that our methodology allows decreasing the uncertainty and the information loss in group decision making and thus, ensures a robust solution to the firm.     

一种区间数-二元联系数转换的模糊决策改进算法

针对准则值和准则权重以二元或三元区间数形式给出的模糊决策问题,提出一种区间数-二元联系数转换改进算法.利用区间数的偏好值和上下限取值范围,将区间数转换为二元联系数.将区间数的偏好值作为联系数的同一度,并将区间数上下限到偏好值的距离作为联系数的差异度,使得转换过程中区间模糊信息中的确定性增大,不确定性减小.在此基础上,使用同一度和差异度重新定义联系数的正负理想解,并确定联系数间的距离公式,进而提出一种改进的基于联系数的TOPSIS模糊决策算法.最后,结合实例表明所提出算法的有效性和合理性.     

An interactive algorithm for multicriteria decision making: theattainable reference point method

An interactive algorithm, the attainable reference point method, is proposed for finding a satisfactory solution to a general multicriteria decision making problem. The decision-maker is only required to modify the reference value of the satisfactory objectives to generate a new attainable reference point in each iteration step. The lexicographic weighted Tchebycheff program associated with the attainable reference point is constructed to guarantee the efficiency of all the discussed points. The value of the unsatisfactory objective chosen by the decision-maker is improved to be satisfactory. Thus its reference value does not need to be modified again in later iterations, and a satisfactory solution can be derived in finite steps. A numerical example is presented to demonstrate the feasibility and efficiency of the proposed method     

An influence-driven feedback system for preference similarity network clustering based consensus group decision making model

Consensus group decision making (CGDM) allows the integration within this area of study of other advanced frameworks such as Social Network Analysis (SNA), Social Influence Network (SIN), clustering and trust-based concepts, among others. These complementary frameworks help to bridge the gap between their corresponding theories in such a way that important elements are not overlooked and are appropriately taken into consideration. In this paper, a new influence-driven feedback mechanism procedure is introduced for a preference similarity network clustering based consensus reaching process. The proposed influence-driven feedback mechanism aims at identifying the network influencer for the generation of advices. This procedure ensures that valuable recommendations are coming from the expert with most similar preferences with the other experts in the group. This is achieved by adapting, from the SIN theory into the CGDM context, an eigenvector-like measure of centrality for the purpose of: (i) measuring the influence score of experts, and (ii) determining the network influencer. Based on the initial evaluations on a set of alternatives provide by the experts in a group, the proposed influence score measure, which is named the σ-centrality, is used to define the similarity social influence network (SSIN) matrix. The σ-centrality is obtained by taking into account both the endogenous (internal network connections) and exogenous (external) factors, which means that SSIN connections as well as the opinion contribution from third parties are permitted in the nomination of the network influencer. The influence-driven feedback mechanism process is designed based on the satisfying of two important conditions to ensure that (1) the revised consensus degree is above the consensus threshold and that (2) the clustering solution is improved.     

An algorithm for characterizing pre-fuzzified linguistic nuance using artificial neural network

The attractiveness of artificial neural networks (ANNs) in solving many complex real world and computational demanding problems was used in characterizing linguistic nuance for harnessing malicious intent or decoding a communication trend. A set of adjectival watch lists was created a priori to serve as target convergence outputs to the ANN’s graphical user interface designed by the researchers. A set of pre-fuzzified or pre-processed speech conversation or written text was used as inputs to the neural network and represents a sub set of actual words used in the investigated two-way communication. The watch lists represents an editable set of words that represents malicious intent or key elements of conversation intent in bidirectional conversation or communication. The watch list database was generated a priori by identifying adjectives and specific nouns as used in the communication under investigation and then normalized. The pre-processed speech and text have been obtained from Recognizers utilizing the hidden Markov models and its hybrids for its processing. The algorithm showed robustness in sorting out pre-normalized and fuzzified speech that ordinarily contained certain elements of interest as conveyed by the investigated conversations. Analysis of a patient-to-healthcare provider’s bidirectional communication during malaria diagnosis and used for testing the developed algorithm showed significant accuracy when compared with the results of clinical analysis or consultation for the corresponding diagnosis.     

An integrated artificial neural network-computer simulation for optimization of complex tandem queue systems

This paper presents an integrated artificial neural network-computer simulation (ANNSim) for optimization of G/G/K queue systems. The ANNSim is a computer program capable of improving its performance by referring to production constraints, system's limitations and desired targets. It is a goal oriented, flexible and integrated approach and produces the optimum solution by utilizing Multi Layer Perceptron (MLP) neural networks. The properties and modules of the prescribed intelligent ANNSim are: (1) parametric modeling, (2) flexibility module, (3) integrated modeling, (4) knowledge-base module, (5) integrated database and (6) learning module. The integrated ANNSim is applied to 30 distinct tandem G/G/K queue systems. Furthermore, its superiority over conventional simulation approach is shown in two dimensions which are average run time and maximum number of required iterations (scenarios).     

An artificial neural network based decision support system for energy efficient ship operations

Reducing fuel consumption of ships against volatile fuel prices and greenhouse gas emissions resulted from international shipping are the challenges that the industry faces today. The potential for fuel savings is possible for new builds, as well as for existing ships through increased energy efficiency measures; technical and operational respectively. The limitations of implementing technical measures increase the potential of operational measures for energy efficient ship operations. Ship owners and operators need to rationalise their energy use and produce energy efficient solutions. Reducing the speed of the ship is the most efficient method in terms of fuel economy and environmental impact. The aim of this paper is twofold: (i) predict ship fuel consumption for various operational conditions through an inexact method, Artificial Neural Network ANN; (ii) develop a decision support system (DSS) employing ANN-based fuel prediction model to be used on-board ships on a real time basis for energy efficient ship operations. The fuel prediction model uses operating data – ‘Noon Data’ – which provides information on a ship’s daily fuel consumption. The parameters considered for fuel prediction are ship speed, revolutions per minute (RPM), mean draft, trim, cargo quantity on board, wind and sea effects, in which output data of ANN is fuel consumption. The performance of the ANN is compared with multiple regression analysis (MR), a widely used surface fitting method, and its superiority is confirmed. The developed DSS is exemplified with two scenarios, and it can be concluded that it has a promising potential to provide strategic approach when ship operators have to make their decisions at an operational level considering both the economic and environmental aspects.