Associate Professor Raymond Chiong

Evolutionary Algorithms (EAs) are population-based, stochastic search algorithms that mimic natural evolution. Due to their ability to find excellent solutions for conventionally hard and dynamic problems within acceptable time, EAs have attracted interest from many researchers and practitioners in recent years. This book "Variants of Evolutionary Algorithms for Real-World Applications" aims to promote the practitioner's view on EAs by providing a comprehensive discussion of how EAs can be adapted to the requirements of various applications in the real-world domains. It comprises 14 chapters, including an introductory chapter re-visiting the fundamental question of what an EA is and other chapters addressing a range of real-world problems such as production process planning, inventory system and supply chain network optimisation, task-based jobs assignment, planning for CNC-based work piece construction, mechanical/ship design tasks that involve runtime-intense simulations, data mining for the prediction of soil properties, automated tissue classification for MRI images, and database query optimisation, among others. These chapters demonstrate how different types of problems can be successfully solved using variants of EAs and how the solution approaches are constructed, in a way that can be understood and reproduced with little prior knowledge on optimisation.

© 2012 Springer-Verlag Berlin Heidelberg. All rights are reserved. The emergence of different metaheuristics and their new variants in recent years has made the definition of the term Evolutionary Algorithms unclear. Originally, it was coined to put a group of stochastic search algorithms that mimic natural evolution together. While some people would still see it as a specific term devoted to this group of algorithms, including Genetic Algorithms, Genetic Programming, Evolution Strategies, Evolutionary Programming, and to a lesser extent Differential Evolution and Estimation of Distribution Algorithms, many others would regard "Evolutionary Algorithms" as a general term describing population-based search methods that involve some form of randomness and selection. In this chapter, we re-visit the fundamental question of "what is an Evolutionary Algorithm?" not only from the traditional viewpoint but also the wider, more modern perspectives relating it to other areas of Evolutionary Computation. To do so, apart from discussing the main characteristics of this family of algorithms we also look at Memetic Algorithms and the Swarm Intelligence algorithms. From our discussion, we see that establishing semantic borders between these algorithm families is not always easy, nor necessarily useful. It is anticipated that they will further converge as the research from these areas cross-fertilizes each other.

The rapid growth of technological advances in recent years has opened a completely new dimension to progress in education and training. The emergence of e-learning has created not only business and educational opportunities, but also significantly improved the standard of the society. This chapter explores the implementation of e-learning and its impact on a community, similar to a university or corporate setup. To this aim, a brief introduction into e-learning technology and an example of using the Blackboard Learning System are brought forth, along with some critical success factors. Projecting the e-learning advantages along with the digital library concepts, the economic benefits of such implementation are highlighted. The discussion then moves to the perspective of students and teachers on e-learning. As the trend in the technological world is moving toward mobility, the wireless e-learning perception is also conferred. In the concluding remarks, e-learning implementation is noted as a positive endeavor to boost economic growth. © 2008, IGI Global.

In earthquake-prone zones, the seismic performance of reinforced concrete cantilever (RCC) retaining walls is a critical factor. In this study, the seismic performance was investigated using horizontal and vertical pseudo-static coefficients. To tackle RCC weights and forces resulting from these earth pressures, 26 constraints for structural strengths and geotechnical stability along with 12 geometric variables are associated with each design. These constraints and design variables form a constraint optimization problem with a 12-dimensional solution space. To conduct effective search and produce sustainable and economical RCC designs that are robust against earthquake hazards, a novel adaptive fuzzy-based metaheuristic algorithm is proposed. The proposed method divides the search space into sub-regions and establishes exploration, information sharing, and exploitation search capabilities based on its novel search components. Further, fuzzy inference systems are employed to address parameterization and computational cost evaluation issues. It was found that the proposed algorithm can achieve low-cost, low-weight, and low-CO2 emission RCC designs under nine seismic conditions when compared with several classical and best-performing design optimizers.

Financial news disclosures provide valuable information for traders and investors while making stock market investment decisions. Essential but challenging, the stock market prediction problem has attracted significant attention from both researchers and practitioners. Conventional machine learning models often fail to interpret the content of financial news due to the complexity and ambiguity of natural language used in the news. Inspired by the success of recurrent neural networks (RNNs) in sequential data processing, we propose an ensemble RNN approach (long short-term memory, gated recurrent unit, and SimpleRNN) to predict stock market movements. To avoid extracting tens of thousands of features using traditional natural language processing methods, we apply sentiment analysis and the sliding window method to extract only the most representative features. Our experimental results confirm the effectiveness of these two methods for feature extraction and show that the proposed ensemble approach is able to outperform other models under comparison.

To address the problem of miss- and false detection during quality inspection of lithium-ion battery cover screen printing (LBCSP), we propose a hybrid image registration method using a point-based feature extraction algorithm and nonlinear-scale space construction. Our proposed method integrates the accelerated-KAZE algorithm with the boosted efficient binary local image descriptor (BEBLID), and is named A-BEBLID. Facing the challenge of the inevitable offset caused by machine vibration during production, we combine a nonlinear diffusion filter with a local image descriptor to extract features from images, and then use the grid-based motion statistics algorithm to remove the incorrect matching pairs. We tested the method on a custom dataset created using images taken from actual lithium-ion battery production lines, named LBCSP. We also evaluated the method on the public HPatches dataset. The average precision achieved by A-BEBLID on the LBCSP dataset is 89% (threshold: 2 pixels), with a localization error of 1.11 pixels, while on the HPatches dataset, the average precision is 73% (threshold: 2 pixels), with a localization error of 1.52 pixels. Comprehensive experimental results also showed that the proposed A-BEBLID can outperform other approaches found in the literature. The method can be further applied to other industry scenarios with similar image registration requirements.

Live stream marketing through social media has attracted the attention of digital retailing marketers in recent years. However, there is a lack of evidence in understanding the influence of in-store live stream on offline purchase intentions. This study aimed to investigate the influence patterns of environmental stimuli on consumers' intention to purchase offline/in-store after watching an in-store live stream session. The Stimuli-Organism-Response (SOR) model was employed as the theoretical framework, and a structured questionnaire was used to collect data from individuals who had previous experience with in-store live stream marketing. Structural equation modelling was then applied for data analysis, with a total of 234 valid responses. The findings revealed that environmental stimuli have a significant positive effect on consumers' intentions to make in-store purchases, and the attitudes towards influencers and products substantially mediate the relationship between stimuli and purchase intention. More specifically, consumer attitude towards products has a pronounced effect on whether they will make an in-store purchase. The novelty of this research lies in its investigation of the impact that live stream marketing has on offline or in-store shopping experiences. This contrasts with the majority of existing live stream studies, which focus on consumers¿ online shopping experiences. In addition, this study broadens the scope of the application of the SOR model to contribute to the growing body of literature on live stream marketing.

Accurate identification of malicious web domains is crucial for protecting users from the risks of theft of private information, malware attack, and monetary loss. Various methods, including blacklists and machine learning-based models, have been proposed to identify malicious web domains effectively. However, maintaining an up-to-date blacklist is difficult, and standard machine learning-based models are typically sensitive to noise in data. In this paper, we propose an ensemble model based on the fuzzy-weighted Least Squares Support Vector Machine (EFW-LS-SVM) for improving malicious web domain identification. Given the fact that different data samples may have varying importance, we introduce a fuzzy-weighted operation by applying it to each data sample. This is the first time the fuzzy-weighted operation has been incorporated into an ensemble approach for malicious web domain identification. Our proposed EFW-LS-SVM delivers excellent results for identifying malicious web domains; it outperformed the compared machine learning models in terms of the F-measure score, as well as provided the best or very competitive accuracy of up to 94.50% for all datasets included in our experiments. Further, considering the imbalanced nature of benign and malicious web domain data, where malicious web domains tend to be the minority, we used the Synthetic Minority Over-sampling Technique (SMOTE) to further improve the performance of all models tested. Our experimental results confirm that SMOTE re-sampling can improve the performance of all the models, including our proposed EFW-LS-SVM¿the F-measure score of EFW-LS-SVM was improved by up to 3.29%.

Increasing energy shortages and environmental pollution have made energy efficiency an urgent concern in manufacturing plants. Most studies looking into sustainable production in general and energy-efficient production scheduling in particular, however, have not paid much attention to logistical factors (e.g., transport and setup). This study integrates multiple automated guided vehicles (AGVs) into a job-shop environment. We propose a multiobjective scheduling model that considers machine processing, sequence-dependent setup and AGV transport, aiming to simultaneously minimize the makespan, total idle time of machines and total energy consumption of both machines and AGVs. To solve this problem, an effective multiobjective evolutionary algorithm (EMOEA) is developed. Within the EMOEA, an efficient encoding/decoding method is designed to represent and decode each solution. A new crossover operator is proposed for AGV assignment and AGV speed sequences. To balance the exploration and exploitation ability of the EMOEA, an opposition-based learning strategy is incorporated. A total of 75 benchmark instances and a real-world case are used for our experimental study. Taguchi analysis is applied to determine the best combination of key parameters for the EMOEA. Extensive computational experiments show that properly increasing the number of AGVs can shorten the waiting time of machines and achieve a balance between economic and environmental objectives for production systems. The experimental results confirm that the proposed EMOEA is significantly better at solving the problem than three other well-known algorithms. Our findings here have significant managerial implications for real-world manufacturing environments integrated with AGVs.

We present an evolutionary trust game to investigate the formation of trust in sharing economy situations, where participants have a fixed provider or consumer role, and can only choose between trustworthy or untrustworthy behaviour. There are a variety of sharing economy platforms catering for differing goods and services, the properties of which may affect the degree to which these roles are variable for users. To the best of our knowledge, this is the first time the evolution of trust in sharing economy situations with fixed provider and consumer roles is being studied in the literature. Our trust model comprises four player types: trustworthy consumer, trustworthy provider, untrustworthy consumer, and untrustworthy provider. Five scenarios with varying initial population ratios of these player types under different host network structures are systematically investigated. Our results show that, in contrast to previous work that allowed switching roles between providers and consumers, trust declines monotonically as the reward for trustworthy behaviour is reduced, with a critical transition point for inversion of trustworthy/untrustworthy populations. In addition, the initial population of trustworthy providers is shown to significantly affect the point at which trustworthy behaviour most declines, with a high proportion resulting in the persistence of trustworthiness even when the reward for such behaviour is significantly low. Our results also show that, when different host network structures are considered, the average degree of the underlying network is an important factor in determining the level of trustworthiness in the population. Our findings may be of importance for understanding the emergence and maintenance of trust in sharing economy platforms where user roles are completely or predominantly rigid.

Machine on/off control is an effective way to achieve energy-efficient production scheduling. Turning off machines and restarting them frequently, however, would incur a considerable amount of additional energy and may even cause damage to the machines. In this paper, we propose a mathematical model based on the energy-efficient flexible job shop scheduling problem (EEFJSP), aiming to minimize not just the makespan and total energy consumption but also the total number of machine restarts. Our idea here is that shifting the start time of operations on different machines appropriately can effectively decrease the number of restarts required and the total energy consumption. We present a two-stage memetic algorithm (TMA) to solve the EEFJSP. A variable neighborhood search approach is designed to improve the convergence speed and fully exploit the solution space of the TMA. An operation-block moving operator is developed to further reduce the total energy consumption as well as the total number of machine restarts without affecting the makespan. Extensive computational experiments carried out to compare the TMA with some well-known algorithms confirm that the proposed TMA can easily obtain better Pareto solutions for the EEFJSP.

There is growing interest in energy-efficient production scheduling research because of the increasing energy shortage. However, most existing studies along this line of research have not considered the energy consumed by automated guided vehicles (AGVs) used in modern smart factories for production scheduling purposes. In this paper, we study an energy-efficient open-shop scheduling problem with multiple AGVs and deteriorating jobs. A multi-objective model with four objectives is formulated, aiming to simultaneously minimise the maximum ending time of all AGVs, the total idle time of machines and AGVs, the total tardiness of jobs, and the total energy consumption of machines and AGVs. An improved population-based multi-objective differential evolution (IMODE) algorithm is developed to solve the problem. The IMODE makes use of a problem feature-based heuristic and a mean entropy method to enhance the diversity of its initial population. A novel grey entropy parallel analysis-based fitness evaluation mechanism with reference points is adopted to evaluate the candidate solutions. To improve the local search ability of IMODE, a multi-level local search strategy is used. In the experimental study, Taguchi analysis is employed to obtain the best parameter combination. The effects of the main components of IMODE are validated via comprehensive comparison experiments. Extensive experimental results show that the IMODE is preferable to other well-known multi-objective algorithms at solving the problem being considered.

In this paper, we propose a novel hybrid fuzzy¿metaheuristic approach with the aim of overcoming premature convergence when solving multimodal single and multi-objective optimization problems. The metaheuristic algorithm used in our proposed approach is based on the imperialist competitive algorithm (ICA), a population-based method for optimization. The ICA divides its population into sub-populations, known as empires. Each empire is composed of a high fitness solution¿the imperialist¿and some lower fitness solutions¿the colonies. Colonies move towards their associated imperialist to achieve better status (higher fitness). The most powerful empire tends to attract weaker colonies. These competitions and movements can be enhanced for better algorithm performance. In our hybrid approach, a global learning strategy is devised for each colony to learn from its best-known position, its associated imperialist and the global best imperialist. A fast-evolutionary elitism local search is used to enhance the collaborative search mechanism (competition) in each empire, and thus the overall optimization performance may be improved. Other main evolutionary operators include velocity adaptation and velocity divergence. To address parameterization and computational cost evaluation issues, two fuzzy inferencing mechanisms are designed and used in parallel: one is a learning strategy adaptor in each run, and the other is a smart evolution selector in each running window. For Pareto front approximation, fast-elitism non-dominated sorting is applied to the solutions, and a novel penalized sigma diversity index is designed to estimate the diversity (power) of solutions in the same rank. Comprehensive experimental results based on 22 single-objective and 25 multi-objective benchmark instances clearly show that our proposed approach provides better solutions compared with other popular metaheuristics and state-of-the-art ICA variants. The proposed approach can be used as an optimization module in any intelligent decision-making systems to enhance efficiency and accuracy. The designed fuzzy inferencing mechanisms can also be incorporated into any single- or multi-objective optimizers for parameter tuning purposes, to make the optimizers more adaptive to new problems or environments.

Mobile-based health (mHealth) systems are proving to be a popular alternative to the traditional visits to healthcare providers. They can also be useful and effective in fighting the spread of infectious diseases, such as the COVID-19 pandemic. Even though young adults are the most prevalent mHealth user group, the relevant literature has overlooked their intention to invest in and use mHealth services. This study aims to investigate the predictors that influence young adults¿ intention to invest in mHealth (IINmH), particularly during the COVID-19 crisis, by designing a research methodology that incorporates both the health belief model (HBM) and the expectation-confirmation model (ECM). As an expansion of the integrated HBM-ECM model, this study proposes two additional predictors: mobile Internet speed and mobile Internet cost. A multi-method analytical approach, including partial least squares structural equation modelling (PLS-SEM), fuzzy-set qualitative comparative analysis (fsQCA), and machine learning (ML), was utilised together with a sample dataset of 558 respondents. The dataset¿about young adults in Bangladesh with an experience of using mHealth¿was obtained through a structured questionnaire to examine the complex causal relationships of the integrated model. The findings from PLS-SEM indicate that value-for-money, mobile Internet cost, health motivation, and confirmation of services all have a substantial impact on young adults¿ IINmH during the COVID-19 pandemic. At the same time, the fsQCA results indicate that a combination of predictors, instead of any individual predictor, had a significant impact on predicting IINmH. Among ML methods, the XGBoost classifier outperformed other classifiers in predicting the IINmH, which was then used to perform sensitivity analysis to determine the relevance of features. We expect this multi-method analytical approach to make a significant contribution to the mHealth domain as well as the broad information systems literature.

Facial expression recognition (FER) is an active research area that has attracted much attention from both academics and practitioners of different fields. In this paper, we investigate an interesting and challenging issue in FER, where the training and testing samples are from a cross-domain dictionary. In this context, the data and feature distribution are inconsistent, and thus most of the existing recognition methods may not perform well. Given this, we propose an effective dynamic constraint representation approach based on cross-domain dictionary learning for expression recognition. The proposed approach aims to dynamically represent testing samples from source and target domains, thereby fully considering the feature elasticity in a cross-domain dictionary. We are therefore able to use the proposed approach to predict class information of unlabeled testing samples. Comprehensive experiments carried out using several public datasets confirm that the proposed approach is superior compared to some state-of-the-art methods.

Obesity is a critical public health problem associated with various complications and diseases. Accurate prediction of body fat is crucial for diagnosing obesity. Various measurement methods, including underwater weighing, dual energy X-ray absorptiometry, bioelectrical impedance analysis, magnetic resonance imaging, air displacement plethysmography, and near infrared interactance, have been used to assess body fat. These measurement methods, however, require special equipment associated with high-cost tests. The aim of this study is to investigate the use of machine learning-based models to accurately predict the body fat percentage. Considering the fact that off-the-shelf machine learning-based models are typically sensitive to noise data, we propose a fuzzy-weighted Gaussian kernel-based Relative Error Support Vector Machine (RE-SVM) for body fat prediction. We first design a fuzzy-weighted operation, which applies fuzzy weights to the error constraints of the RE-SVM, to alleviate the influence of noise data. Next, we also apply the fuzzy weights to improve the Gaussian kernel by considering the importance of different samples. Computational experiments and statistical tests conducted confirm that our proposed approach is able to significantly outperform other models being compared for body fat prediction across different performance metrics used. The proposed approach offers a viable alternative for diagnosing obesity when high-cost measurement methods are not available.

Scholarship on displacement caused by the effects of climate change generally approaches displacement as the involuntary movement of people. However, in this article, we argue that there are uncertainties surrounding Climate Change Induced Displacement (CCID) that are partly caused by discursive ambiguity around the notion of ¿displacement¿¿a concept that remains poorly defined in the context of climate change research¿and a conflation between displacement due to quick-onset disaster events and the cumulative pressure of living in an environment marked by a disrupted climate. Reflecting on the impacts of the Australian bushfires in 2019¿20, we conceptualise CCID beyond migration as an event and a physical relocation across geographical space. Even fast-onset disaster events, such as the Australian bushfires, can dispossess and displace beyond the immediate threat of the fire front; but this displacement is not necessarily aligned with movement and migration, nor is it evenly proportioned across populations. Based on a review of existing literature on CCID, we identify three key tensions shaping scholarship on CCID: conceptualisation; distribution of risk and impact; and discursive framing. Together, we contend, these tensions highlight the imperative of striving for conceptual clarity and awareness of distributional inequities of risk and vulnerabilities.

This article presents a new model to handle the cast break problem caused by small daily disruptions in the processing time of the steelmaking and continuous casting (SCC) production process. In this model, the exact distribution of the uncertain parameters is unknown, and support set, mean, and covariance information is used to describe the uncertain processing time. The problem aims to determine the assignments, sequences, and time points of the charges to be processed on corresponding machines. The main goal is to minimize the expected value of the production objective while reducing the number of cast break occurrences. The problem is solved in two steps. First, a subproblem is developed by fixing the sequences and the assignments of the charges. This subproblem is formulated as a distributionally robust chance-constrained (DRCC) model, in which the constraints are established with certain probabilities even when the uncertain processing times are in their worst cases. A dual approximation method is proposed to convert the model into a semidefinite programming problem so that it can be solved by standard solvers. Additionally, a linear programming approximation method is used to accelerate the solving procedure. A Tabu search algorithm incorporated with a speed-up strategy is also designed to determine the assignments and sequences of the charges. Both simulated data generated from different distributions and actual production data are used to test the efficacy of our model. Results of the numerical experiments show that the schedule obtained from the DRCC model is more robust, i.e., it causes fewer cast breaks than the nominal schedule obtained from a deterministic model.

Background: Alzheimer's disease (AD) is one of the deadliest diseases in developed countries. Treatments following early AD detection can significantly delay institutionalisation and extend patients' independence. There has been a growing focus on early AD detection using artificial intelligence. Convolutional neural networks (CNNs) have proven revolutionary for image-based applications and have been applied to brain scans. In recent years, studies have utilised two-dimensional (2D) CNNs on magnetic resonance imaging (MRI) scans for AD detection. To apply a 2D CNN on three-dimensional (3D) MRI volumes, each MRI scan is split into 2D image slices. A CNN is trained over the image slices by calculating a loss function between each subject's label and each image slice's predicted output. Although 2D CNNs can discover spatial dependencies in an image slice, they cannot understand the temporal dependencies among 2D image slices in a 3D MRI volume. This study aims to resolve this issue by modelling the sequence of MRI features produced by a CNN with deep sequence-based networks for AD detection. Method: The CNN utilised in this paper was ResNet-18 pre-trained on an ImageNet dataset. The employed sequence-based models were the temporal convolutional network (TCN) and different types of recurrent neural networks. Several deep sequence-based models and configurations were implemented and compared for AD detection. Results: Our proposed TCN model achieved the best classification performance with 91.78% accuracy, 91.56% sensitivity and 92% specificity. Conclusion: Our results show that applying sequence-based models can improve the classification accuracy of 2D and 3D CNNs for AD detection by up to 10%.

New energy vehicles (NEVs) are welcomed by both policymakers and consumers because of their energy saving and low carbon properties. However, due to their high production cost and limited cruising range, the development of the NEV industry relies heavily on governments¿ cash subsidy (CS) programs. At the same time, policymakers in several countries, including the United States (California) and China, have introduced carbon regulation (CR) to re-energize the NEV market. CS and CR have different impacts on the consumer and production sides. To this end, we propose a hybrid model combining the advantages of both. Through this model, the impacts of different interventions on product demands, profits, carbon emissions and technology R&D are derived: (1) CS always increases the demand for NEVs, while CR promotes NEVs only when NEVs are emission saving; providing CS increases the profits of the supply chain members, but supply chain profits under CR show uncertain trends; (2) CS reduces total emissions only if the unit carbon emission of NEVs is small enough, and the effect of NEVs¿ carbon quota on total emissions under CR shows a similar trend, whereas the impact of carbon price on total emissions under CR depends on specific thresholds. Further results show that when a customer values NEVs highly, the proposed hybrid model produces the lowest total emission; (3) Optimal technology R&D for the cruising range of NEVs increases under CS if the subsidy provided is based on the actual cruising range; otherwise, CR performs better. The hybrid model shows potential in outperforming the other two policies in terms of the optimal technology R& D; (4) Supply chain integration decreases the optimal technology R&D in the case of CR, whereas in the case of CS, technology R&D does not vary with integration or decentralization of the supply chain.

Numerous studies on mental depression have found that tweets posted by users with major depressive disorder could be utilized for depression detection. The potential of sentiment analysis for detecting depression through an analysis of social media messages has brought increasing attention to this field. In this article, we propose 90 unique features as input to a machine learning classifier framework for detecting depression using social media texts. Derived from a combination of feature extraction approaches using sentiment lexicons and textual contents, these features are able to provide impressive results in terms of depression detection. While the performance of different feature groups varied, the combination of all features resulted in accuracies greater than 96% for all standard single classifiers, and the best accuracy of over 98% with Gradient Boosting, an ensemble classifier.

Pt is a key high-performing catalyst for important chemical conversions, such as biomass conversion and water splitting. Limited Pt reserves, however, demand that we identify more sustainable alternative catalyst materials for these processes. Here, we combine state-of-the-art graph neural networks and crystal graph machine learning representations with active learning to discover new, low-cost Pt alloy catalysts for biomass reforming and hydrogen evolution reactions. We identify 12 Pt-based alloys which have comparable catalytic activity to that of the exemplar Pt(111) surface. Notably, Cu3Pt and FeCuPt2 exhibit near identical catalytic performance as that of Pt(111). These results demonstrate the potential of machine learning for predicting new catalytic materials without recourse to expensive DFT geometry optimizations, the current bottleneck impeding high-throughput materials discovery. We also examine the performance of d-band theory in elucidating trends in binary and ternary Pt alloys.

The financial impact of positive reviews has prompted some fraudulent sellers to generate fake product reviews for either promoting their products or discrediting competing products. Many e-commerce portals have implemented measures to detect such fake reviews, and these measures require excellent detectors to be effective. In this work, we propose 133 unique features from the combination of content and behaviour-based features to detect fake reviews using machine learning classifiers. Preliminary results show that these features can provide good results for all datasets tested. Detailed analysis of the results, however, reveals the existence of class imbalance issues for two of the bigger datasets - there is a high imbalance between the accuracies of different classes (e.g., 7.73% for the fake class and 99.3% for the genuine class using a Multilayer Perceptron classifier). We therefore introduce two sampling methods that can improve the accuracy of the fake review class on balanced datasets. The accuracies can be improved to a maximum of 89% for both random under and over-sampling on Convolutional Neural Networks. Additionally, we propose a parallel cross-validation method that can speed up the validation process in a parallel environment.

This paper presents a computational evolutionary game model to study and understand fraud dynamics in the consumption tax system. Players are cooperators if they correctly declare their value added tax (VAT), and are defectors otherwise. Each player's payoff is influenced by the amount evaded and the subjective probability of being inspected by tax authorities. Since transactions between companies must be declared by both the buyer and seller, a strategy adopted by one influences the other?s payoff. We study the model with a wellmixed population and different scalefree networks. Model parameters were calibrated using real-world data of VAT declarations by businesses registered in the Canary Islands region of Spain. We analyzed several scenarios of audit probabilities for high and low transactions and their prevalence in the population, as well as social rewards and penalties to find the most efficient policy to increase the proportion of cooperators. Two major insights were found. First, increasing the subjective audit probability for low transactions is more efficient than increasing this probability for high transactions. Second, favoring social rewards for cooperators or alternative penalties for defectors can be effective policies, but their success depends on the distribution of the audit probability for low and high transactions.

Recent large-scale migration flows from rural areas of the Mekong Delta (MKD) to larger cities in the South-East (SE) region of Vietnam have created the largest migration corridor in the country. This migration trend has further contributed to greater rural¿urban disparities and widened the development gap between regions. In this study, our aim is to understand the migration dynamics and determine the most critical factors affecting the behavior of migrants in the MKD region. We present an agent-based model and incorporate the Theory of Planned Behavior to effectively break down migration intention into related components and contributing factors. A genetic algorithm is used for automated calibration and sensitivity analysis of model parameters, in order to validate our agent-based model. We further explore the migration behavior of people in certain demographic groups and delineate migration flows across cities and provinces from the MKD to the SE region.

External factors, such as the increasing environmental awareness among consumers and introduction of environmental regulations by governments, have stimulated manufacturers to produce new green products. Cost factors, on the other hand, encourage the continuation of older generation products and hinder the launch of new green products. To study this dilemma, we consider a single manufacturer with the ability of cost learning from the production of an older generation product, and it intends to launch a new green product. We first derive the optimal pricing and production strategies of the manufacturer based on accumulated cost learning using dynamic programming techniques. Then, we identify the threshold conditions for producing the new green product and the Pareto area that can increase both the profit of the manufacturer and surplus of consumers. For comparison purposes, we also study the impact of external regulations such as environmental taxes on the introduction of a new green product ¿ with results suggesting that the implementation of environmental taxes promotes the accessibility of green products and reduces the Pareto area of profit and consumer surplus. We also find that, under the influence of cost learning, the introduction of new greener products and the implementation of environmental taxes serve to reduce the total environmental damage done by the manufacturer. Results of our models should provide instructive managerial insights for the introduction of new greener products.

This paper studies an important extension of energy-efficient production scheduling research, where machine on/off control and machine maintenance are considered simultaneously. The inspiration of this extension is that a machine must be turned off if it needs to be maintained, and an already-turned-off machine can be maintained without needing to be restarted. We therefore formulate an energy-efficient production scheduling problem with machine maintenance through machine on/off control, aiming to optimise three objectives ¿ the makespan, total number of machine restarts, and energy consumption ¿ at the same time. Four rules are designed to set the machine on/off criteria, maintenance periods and predefined maintenance windows, based on solutions of the job shop scheduling problem (JSP) as a test case. Three heuristics are proposed to insert the maintenance activities into the solutions and move their maintenance-operation blocks to optimise the objectives. The effectiveness of the first rule and the moving of maintenance-operation blocks have been proven mathematically. Our proposed heuristics, unlike traditional heuristic algorithms, are expected to be applicable and effective even if we change the objectives and constraints, require minimal computational time (only a few seconds) to optimise a scheduling solution, and can solve different types of scheduling problems without needing any modification. Experiments undertaken indicate promising performance of the proposed heuristics based on 182 JSP benchmark instances.

Online reviews are becoming increasingly important for decision-making. Consumers often refer to online reviews for opinions before making a purchase. Marketers also acknowledge the importance of online reviews and use them to improve product success. However, the massive amount of online review data, as well as its unstructured nature, is a challenge for anyone wanting to derive a conclusion quickly. In this paper, we propose a novel framework for gauging the ratings of online reviews using machine learning techniques. This framework uses a combination of text pre-processing and feature extraction methods. Here, we investigate four different aspects of the new framework. First, we assess the performance of single and ensemble classifiers in predicting sentiment¿positive or negative¿initially on a specific dataset (Yelp), but subsequently also on two other datasets (Amazon's product reviews and a¿movie review dataset). Second, using the best identified classifiers, we improve the accuracy with which neutral polarity can be predicted, an ability largely overlooked in the literature. Third, we further improve the performance of these classifiers by testing different pre-processing and feature extraction methods. Finally, we measure how well our deep learning approach performs on the same task compared to the best previously identified classifiers. Our extensive testing shows that the linear-kernel support vector machine, logistic regression and multilayer perceptron are the three¿best single classifiers in terms of accuracy, precision, recall, and F-measure. Their performance could be further improved if they were used as base classifiers for ensemble models. We also observe that several text pre-processing techniques¿negation word identification, word elongation correction, and part of speech lemmatisation (combined with Terms Frequency¿and N-gram¿words)¿can increase accuracy. In addition, we demonstrate that the general sentiment of lexicons such as SentiWordNet¿3.0 and SenticNet 4 can be used to generate features with good results, although deep learning models can perform equally well. Experiments with different datasets confirm that our framework provides consistent outcomes. In particular, we have focused on improving the accuracy of neutral sentiment, and we conclude by showing how this can be achieved without sacrificing the accuracy of positive or negative ratings.

Depression is one of the leading causes of suicide worldwide. However, a large percentage of cases of depression go undiagnosed and, thus, untreated. Previous studies have found that messages posted by individuals with major depressive disorder on social media platforms can be analysed to predict if they are suffering, or likely to suffer, from depression. This study aims to determine whether machine learning could be effectively used to detect signs of depression in social media users by analysing their social media posts¿especially when those messages do not explicitly contain specific keywords such as ¿depression¿ or ¿diagnosis¿. To this end, we investigate several text preprocessing and textual-based featuring methods along with machine learning classifiers, including single and ensemble models, to propose a generalised approach for depression detection using social media texts. We first use two public, labelled Twitter datasets to train and test the machine learning models, and then another three non-Twitter depression-class-only datasets (sourced from Facebook, Reddit, and an electronic diary) to test the performance of our trained models against other social media sources. Experimental results indicate that the proposed approach is able to effectively detect depression via social media texts even when the training datasets do not contain specific keywords (such as ¿depression¿ and ¿diagnose¿), as well as when unrelated datasets are used for testing.

This paper presents an effective multi-objective Jaya (EMOJaya) algorithm to solve a multi-objective job-shop scheduling problem, aiming to simultaneously minimise the makespan, total flow time and mean tardiness. A strategy based on grey entropy parallel analysis (GEPA) is developed to assess and select solutions during the search process. To obtain a high-quality reference sequence for GEPA, an opposition-based learning (OBL) strategy is used in parallel. Additionally, the OBL strategy is incorporated into Jaya's search operation and external archive to enhance the search ability and convergence rate of the algorithm. Computational experiments based on 30 benchmark instances with different scales confirm that GEPA and OBL can significantly improve the performance of our proposed EMOJaya. Experimental results also show that EMOJaya is able to outperform three state-of-the-art multi-objective algorithms in solving the problem at hand in terms of convergence, diversity and distribution. Further, EMOJaya can obtain more high-quality scheduling schemes, which provide more and better options for decision makers.

Optimization techniques have drawn much attention for solving geotechnical engineering problems in recent years. Particle swarm optimization (PSO) is one of the most widely used population-based optimizers with a wide range of applications. In this paper, we first provide a detailed review of applications of PSO on different geotechnical problems. Then, we present a comprehensive computational study using several variants of PSO to solve three specific geotechnical engineering benchmark problems: the retaining wall, shallow footing, and slope stability. Through the computational study, we aim to better understand the algorithm behavior, in particular on how to balance exploratory and exploitative mechanisms in these PSO variants. Experimental results show that, although there is no universal strategy to enhance the performance of PSO for all the problems tackled, accuracies for most of the PSO variants are significantly higher compared to the original PSO in a majority of cases.

Dyeing is the most time and energy-consuming process in textile production. Motivated by a dyeing overdue problem in a lace textile factory, we study a parallel machine scheduling problem with different colour families, sequence-dependent setup times, and machine eligibility restriction. An integer programming model is formulated to minimise the total tardiness. Given that the dyeing optimisation problem is strongly NP-hard, a hybrid differential evolution (HDE) algorithm embedded with chaos theory and two local search algorithms is proposed to solve real-world instances from the textile factory. In our proposed algorithm, a special encoding and decoding scheme is designed to deal with the machine eligibility constraint, and chaos theory is adopted to determine the parameter settings of the underlying differential evolution (DE) algorithm. To speed up convergence and improve search exploitation, two local search algorithms inspired by two dominance properties are developed to determine the optimal job sequence for parallel machines, such that the decision of the entire problem is simplified to the assignment of jobs among the machines, and the computational time required is significantly reduced. Comprehensive experiments based on 36 synthetically generated small to large-scale problem instances and 20 real-world industrial data sets confirm the efficacy of our proposed HDE over other DE variants.

Fraudulent online sellers often collude with reviewers to garner fake reviews for their products. This act undermines the trust of buyers in product reviews, and potentially reduces the effectiveness of online markets. Being able to accurately detect fake reviews is, therefore, critical. In this study, we investigate several preprocessing and textual-based featuring methods along with machine learning classifiers, including single and ensemble models, to build a fake review detection system. Given the nature of product review data, where the number of fake reviews is far less than that of genuine reviews, we look into the results of each class in detail in addition to the overall results. We recognise from our preliminary analysis that, owing to imbalanced data, there is a high imbalance between the accuracies for different classes (e.g., 1.3% for the fake review class and 99.7% for the genuine review class), despite the overall accuracy looking promising (around 89.7%). We propose two dynamic random sampling techniques that are possible for textual-based featuring methods to solve this class imbalance problem. Our results indicate that both sampling techniques can improve the accuracy of the fake review class¿for balanced datasets, the accuracies can be improved to a maximum of 84.5% and 75.6% for random under and over-sampling, respectively. However, the accuracies for genuine reviews decrease to 75% and 58.8% for random under and over-sampling, respectively. We also discover that, for smaller datasets, the Adaptive Boosting ensemble model outperforms other single classifiers; whereas, for larger datasets, the performance improvement from ensemble models is insignificant compared to the best results obtained by single classifiers.

Purpose: The aim of this study is to propose an efficient rule extraction and integration approach for identifying phishing websites. The proposed approach can elucidate patterns of phishing websites and identify them accurately. Design/methodology/approach: Hyperlink indicators along with URL-based features are used to build the identification model. In the proposed approach, very simple rules are first extracted based on individual features to provide meaningful and easy-to-understand rules. Then, the F-measure score is used to select high-quality rules for identifying phishing websites. To construct a reliable and promising phishing website identification model, the selected rules are integrated using a simple neural network model. Findings: Experiments conducted using self-collected and benchmark data sets show that the proposed approach outperforms 16 commonly used classifiers (including seven non¿rule-based and four rule-based classifiers as well as five deep learning models) in terms of interpretability and identification performance. Originality/value: Investigating patterns of phishing websites based on hyperlink indicators using the efficient rule-based approach is innovative. It is not only helpful for identifying phishing websites, but also beneficial for extracting simple and understandable rules.

This paper presents a distributionally robust (DR) optimization model for the single machine scheduling problem (SMSP) with random job processing time (JPT). To the best of our knowledge, it is the first time a DR optimization approach is applied to production scheduling problems in the literature. Unlike traditional stochastic programming models, which require an exact distribution, the presented DR-SMSP model needs only the mean-covariance information of JPT. Its aim is to find an optimal job sequence by minimizing the worst-case Conditional Value-at-Risk (Robust CVaR) of the job sequence's total flow time. We give an explicit expression of Robust CVaR, and decompose the DR-SMSP into an assignment problem and an integer second-order cone programming (I-SOCP) problem. To efficiently solve the I-SOCP problem with uncorrelated JPT, we propose three novel Cauchy-relaxation algorithms. The effectiveness and efficiency of these algorithms are evaluated by comparing them to a CPLEX solver, and robustness of the optimal job sequence is verified via comprehensive simulation experiments. In addition, the impact of confidence levels of CVaR on the tradeoff between optimality and robustness is investigated from both theoretical and practical perspectives. Our results convincingly show that the DR-SMSP model is able to enhance the robustness of the optimal job sequence and achieve risk reduction with a small sacrifice on the optimality of the mean value. Through the simulation experiments, we have also been able to identify the strength of each of the proposed algorithms.

The industrial sector is one of the largest energy consumers in the world. To alleviate the grid's burden during peak hours, time-of-use (TOU) electricity pricing has been implemented in many countries around the globe to encourage manufacturers to shift their electricity usage from peak periods to off-peak periods. In this paper, we study the unrelated parallel machine scheduling problem under a TOU pricing scheme. The objective is to minimize the total electricity cost by appropriately scheduling the jobs such that the overall completion time does not exceed a predetermined production deadline. To solve this problem, two solution approaches are presented. The first approach models the problem with a new time-interval-based mixed integer linear programming formulation. In the second approach, we reformulate the problem using Dantzig-Wolfe decomposition and propose a column generation heuristic to solve it. Computational experiments are conducted under different TOU settings and the results confirm the effectiveness of the proposed methods. Based on the numerical results, we provide some practical suggestions for decision makers to help them in achieving a good balance between the productivity objective and the energy cost objective.

Due to the huge volume and linguistic variation of data shared online, accurate detection of the sentiment of a message (polarity detection) can no longer rely on human assessors or through simple lexicon keyword matching. This paper presents a semi-supervised approach in constructing essential toolkits for analysing the polarity of a localised scarce-resource language, Singlish (Singaporean English). Corpus-based bootstrapping using a multilingual, multifaceted lexicon was applied to construct an annotated testing dataset, while unsupervised methods such as lexicon polarity detection, frequent item extraction through association rules and latent semantic analysis were used to identify the polarity of Singlish n-grams before human assessment was done to isolate misleading terms and remove concept ambiguity. The findings suggest that this multilingual approach outshines polarity analysis using only the English language. In addition, a hybrid combination of the Support Vector Machine and a proposed Singlish Polarity Detection algorithm, which incorporates unigram and n-gram Singlish sentic patterns with other multilingual polarity sentic patterns such as negation and adversative, is able to outperform other approaches in comparison. The promising results of a pooled testing dataset generated from the vast amount of unannotated Singlish data clearly show that our multilingual Singlish sentic pattern approach has the potential to be adopted in real-world polarity detection.

In recent years, there has been a growing concern over the environmental impact of traditional manufacturing, especially in terms of energy consumption and related emissions of carbon dioxide. Besides the adoption of new equipment, production scheduling could play a key role in reducing the total energy consumption of a manufacturing plant. In this paper, we explicitly introduce the objective of minimizing energy consumption into a typical production scheduling model, i.e., the job shop scheduling problem, based on a machine speed scaling framework. To solve this bi-objective optimization problem, we propose a multi-objective genetic algorithm incorporated with two problem-specific local improvement strategies. These local improvement procedures aim to enhance the solution quality by utilizing the mathematical models of two restricted subproblems derived from the original problem. Comprehensive computational experiments have been carried out to verify the effectiveness of the proposed solution approach. The results presented in this work may be useful for future research on energy-efficient production scheduling.

Local search algorithms such as Ejection Chain Methods (ECMs) based on the stem-and-cycle (S&C) reference structure, Lin-Kernighan (LK) heuristics, Tabu Search (TS) as well as the recently proposed Multi-Neighborhood Search (MNS) have been found to be highly competitive for solving the Traveling Salesman Problem (TSP). In this paper, we carry out a large-scale experimental study with all 110 symmetric instances from the TSPLib to investigate the performance of these algorithms. Our study is different from previous work along this line of research in that we consider the entire runtime behavior of the algorithms rather than just their end results. This leads to one of the most comprehensive comparisons of these algorithms using the TSP instances. We then introduce an improved S&C-ECM (named FSM**) that can outperform LK, TS, and MNS. In order to further boost the performance, we develop new hybrid versions of our ECM implementations by combining them with Evolutionary Algorithms and Population-based Ant Colony Optimization. We compare them to similar hybrids of LK, TS, and MNS. Our results show that hybrid algorithms of S&C-ECM, LK, TS and MNS are all very efficient for solving the TSP. We also find that the full runtime behavior comparison provides deeper and clearer insights, while focusing on end results only could have led to a misleading conclusion.

We investigated the critical success factors that affect the implementation of electronic document management systems in government organizations. Based on a comprehensive literature review and input from an expert panel, we composed a list of 37 factors that were considered as prerequisites of successful electronic document management system implementation. We then grouped these 37 factors into six categories. Through a questionnaire survey and factor analysis, we confirmed that the categories identified are important for successful electronic document management system implementation.

Interval time series prediction is one of the most challenging research topics in the field of time series modeling and prediction. In view of the remarkable function approximation capability of fully complex-valued radial basis function neural networks (FCRBFNNs), we set out to investigate the possibility of forecasting interval time series by denoting the lower and upper bounds of the interval as real and imaginary parts of a complex number, respectively. This results in a complex-valued interval. We then model the resulted complex-valued interval time series via a FCRBFNN. Furthermore, we propose to evolve the FCRBFNN by using particle swarm optimization (PSO) and discrete PSO for joint optimization of the structure and parameters. Finally, the proposed interval time series prediction approach is tested with simulated interval time series data as well as real interval stock price time series data from the New York Stock Exchange. Our experimental results indicate that it is a promising alternative for interval time series forecasting.

Selection of input features plays an important role in developing models for short-term load forecasting (STLF). Previous studies along this line of research have focused pre-dominantly on filter and wrapper methods. Given the potential value of a hybrid selection scheme that includes both filter and wrapper methods in constructing an appropriate pool of features, coupled with the general lack of success in employing filter or wrapper methods individually, in this study we propose a hybrid filter-wrapper approach for STLF feature selection. This proposed approach, which is believed to have taken full advantage of the strengths of both filter and wrapper methods, first uses the Partial Mutual Information based filter method to filter out most of the irrelevant and redundant features, and subsequently applies a wrapper method, implemented via a firefly algorithm, to further reduce the redundant features without degrading the forecasting accuracy. The well-established support vector regression is selected as the modeler to implement the proposed hybrid feature selection scheme. Real-world electricity load datasets from a North-American electric utility and the Global Energy Forecasting Competition 2012 have been used to test the performance of the proposed approach, and the experimental results show its superiority over selected counterparts.

This paper presents a novel generalised indirect reciprocity approach for promoting cooperation in social dilemma games. Here, players decide upon an action to play in the game based on public information (or ¿external cues¿) rather than individual-specific information. The public information is constantly updated according to the underlying learning model. Comprehensive simulation experiments using the N-player Prisoner's Dilemma (PD) and Snowdrift (SD) games show that generalised indirect reciprocity promotes high levels of cooperation across a wide range of conditions. This is despite the fact that the make-up of player groups is continually changing. As expected, the extent of cooperative behaviour observed in the ¿constraint-relaxed¿ N-player SD game is significantly higher than the N-player PD game. Our proposed generalised indirect reciprocity model may shed light on the conundrum of cooperation between anonymous individuals.

This paper proposes a Tabu-mechanism improved iterated greedy (TMIIG) algorithm to solve the no-wait flowshop scheduling problem with a makespan criterion. The idea of seeking further improvement in the iterated greedy (IG) algorithm framework is based on the observation that the construction phase of the original IG algorithm may not achieve good performance in escaping from local minima when incorporating the insertion neighborhood search. To overcome this limitation, we have modified the IG algorithm by utilizing a Tabu-based reconstruction strategy to enhance its exploration ability. A powerful neighborhood search method that involves insert, swap, and double-insert moves is then applied to obtain better solutions from the reconstructed solution in the previous step. Empirical results on several benchmark problem instances and those generated randomly confirm the advantages of utilizing the new reconstruction scheme. In addition, our results also show that the proposed TMIIG algorithm is relatively more effective in minimizing the makespan than other existing well-performing heuristic algorithms.

Protein structure prediction is a fundamental issue in the field of computational molecular biology. In this paper, the AB off-lattice model is adopted to transform the original protein structure prediction scheme into a numerical optimization problem. We present a balance-evolution artificial bee colony (BE-ABC) algorithm to address the problem, with the aim of finding the structure for a given protein sequence with the minimal free-energy value. This is achieved through the use of convergence information during the optimization process to adaptively manipulate the search intensity. Besides that, an overall degradation procedure is introduced as part of the BE-ABC algorithm to prevent premature convergence. Comprehensive simulation experiments based on the well-known artificial Fibonacci sequence set and several real sequences from the database of Protein Data Bank have been carried out to compare the performance of BE-ABC against other algorithms. Our numerical results show that the BE-ABC algorithm is able to outperform many state-of-the-art approaches and can be effectively employed for protein structure optimization.

Stochastic approaches such as evolutionary algorithms have been widely used in various science and engineering problems. When comparing the performance of a set of stochastic algorithms, it is necessary to statistically evaluate which algorithms are the most suitable for solving a given problem. The outcome of statistical tests comparing N=2 processes, where N is the number of algorithms, is often presented in tables. This can become confusing for larger numbers of N. Such a scenario is, however, very common in both numerical and combinatorial optimization as well as in the domain of stochastic algorithms in general. In this letter, we introduce an alternative way of visually presenting the results of statistical tests for multiple processes in a compact and easy-to-read manner using a directed acyclic graph (DAG), in the form of a simplified Hasse diagram. The rationale of doing so is based on the fact that the outcome of the tests is always at least a strict partial order, which can be appropriately presented via a DAG. The goal of this brief communication is to promote the use of this approach as a means for presenting the results of comparisons between different optimization methods.

Accurate forecasting of mid-term electricity load is an important issue for power system planning and operation. Instead of point load forecasting, this study aims to model and forecast mid-term interval loads up to one month in the form of interval-valued series consisting of both peak and valley points by using MSVR (Multi-output Support Vector Regression). In addition, an MA (Memetic Algorithm) based on the firefly algorithm is used to select proper input features among the feature candidates, which include time lagged loads as well as temperatures. The capability of this proposed interval load modeling and forecasting framework to predict daily interval electricity demands is tested through simulation experiments using real-world data from North America and Australia. Quantitative and comprehensive assessments are performed and the experimental results show that the proposed MSVR-MA forecasting framework may be a promising alternative for interval load forecasting.

© 2015 Elsevier B.V. Electronic government (or e-government) initiatives are widespread across the globe. The increasing interest in e-government raises the issue of how governments can increase citizen adoption and usage of their online services. In this study, the fundamental argument is that citizens can be viewed as customers, and that e-government success can be measured by the extent to which customer net benefits are positively influenced. Hence, the key consequents of e-government success are customer-related, and the antecedents of such success have to be considered from the customer viewpoint. We advocate that government agencies must consider their customers' perceptions of empowerment as a key causal mechanism in deriving value from e-government systems. However, the literature appears to lack this perspective. This study aims to fill the gap by proposing a theoretical model and an associated evaluation tool that measures the e-government performance from a customer empowerment perspective. The model was validated by a survey method and analyzed using partial least squares. The results support our argument and show that all paths in the proposed model are significant.

Background: DNA microarray gene expression classification poses a challenging task to the machine learning domain. Typically, the dimensionality of gene expression data sets could go from several thousands to over 10,000 genes. A potential solution to this issue is using feature selection to reduce the dimensionality. Aim The aim of this paper is to investigate how we can use feature quality information to improve the precision of microarray gene expression classification tasks. Method: We propose two evolutionary machine learning models based on the eXtended Classifier System (XCS) and a typical feature selection methodology. The first one, which we call FS-XCS, uses feature selection for feature reduction purposes. The second model is GRD-XCS, which uses feature ranking to bias the rule discovery process of XCS. Results: The results indicate that the use of feature selection/ranking methods is essential for tackling high-dimensional classification tasks, such as microarray gene expression classification. However, the results also suggest that using feature ranking to bias the rule discovery process performs significantly better than using the feature reduction method. In other words, using feature quality information to develop a smarter learning procedure is more efficient than reducing the feature set. Conclusion: Our findings have shown that extracting feature quality information can assist the learning process and improve classification accuracy. On the other hand, relying exclusively on the feature quality information might potentially decrease the classification performance (e.g., using feature reduction). Therefore, we recommend a hybrid approach that uses feature quality information to direct the learning process by highlighting the more informative features, but at the same time not restricting the learning process to explore other features.

This paper focuses on detecting Alzheimer's Disease (AD) using the ResNet-18 model on Magnetic Resonance Imaging (MRI). Previous studies have applied different 2D Convolutional Neural Networks (CNNs) to detect AD. The main idea being to split 3D MRI scans into 2D image slices, so that classification can be performed on the image slices independently. This idea allows researchers to benefit from the concept of transfer learning. However, 2D CNNs are incapable of understanding the relationship among 2D image slices in a 3D MRI scan. One solution is to employ 3D CNNs instead of 2D ones. In this paper, we propose a method to utilise transfer learning in 3D CNNs, which allows the transfer of knowledge from 2D image datasets to a 3D image dataset. Both 2D and 3D CNNs are compared in this study, and our results show that introducing transfer learning to a 3D CNN improves the accuracy of an AD detection system. After using an optimisation method in the training process, our approach achieved 96.88% accuracy, 100% sensitivity, and 93.75% specificity.

Local search such as Ejection Chain Methods (ECMs) based on the stem-and-cycle (S&C) reference structure, Lin-Kernighan (LK) heuristics, as well as the recently proposed Multi-Neighborhood Search (MNS), are among the most competitive algorithms for the Traveling Salesman Problem (TSP). In this paper, we carry out a large-scale experiment with all 110 symmetric instances from the TSPLib to investigate the performances of these algorithms. Our study is different from previous work along this line of research in that we consider the entire runtime behavior of the algorithms, not just their end results. This leads to one of the most comprehensive comparisons of these algorithms to date. We introduce a new, improved S&C-ECM that can outperform LK and MNS. We then develop new hybrid versions of our ECM implementations by combining them with Evolutionary Algorithms and Population-based Ant Colony Optimization (PACO). We compare them to similar hybrids of LK and MNS. Our results show that hybrid PACO-S&C, PACO-LK and PACO-MNS are all very efficient. We also find that the full runtime behavior comparison provides deeper and clearer insights, while focusing on end results only would have led to a misleading conclusion.

With the large and growing user base of social media, it is not an easy feat to identify potential customers for business. This is mainly due to the challenge of extracting commercially viable contents from the vast amount of free-form conversations. In this paper, we analyse the Twitter content of an account owner and its list of followers through various text mining methods and segment the list of followers via an index. We have termed this index as the High-Value Social Audience (HVSA) index. This HVSA index enables a company or organisation to devise their marketing and engagement plan according to available resources, so that a high-value social audience can potentially be transformed to customers, and hence improve the return on investment.

Submarine search-evasion path planning aims to acquire an evading route for a submarine so as to avoid the detection of hostile anti-submarine searchers such as helicopters, aircraft and surface ships. In this paper, we propose a numerical optimization model of search-evasion path planning for invading submarines. We use the Artificial Bee Colony (ABC) algorithm, which has been confirmed to be competitive compared to many other nature-inspired algorithms, to solve this numerical optimization problem. In this work, several search-evasion cases in the two-dimensional plane have been carefully studied, in which the anti-submarine vehicles are equipped with sensors with circular footprints that allow them to detect invading submarines within certain radii. An invading submarine is assumed to be able to acquire the real-time locations of all the anti-submarine searchers in the combat field. Our simulation results show the efficacy of our proposed dynamic route optimization model for the submarine search-evasion path planning mission.

This paper presents a method on modelling the economy market using agent-based representation and iterated prisoner's dilemma (IPD). While IPD has been widely used in various problems, there is little work done in modelling the dynamics of market behaviour and it is the main objective of this work to fill in this gap. We create a simulated market environment with agents acting as firms to perform transactions among each other with chosen IPD strategy. From the empirical results, we investigate the strategic interactions among different firms in relation to real-world transaction scenario. In the concluding remarks, we present our observation on the qualities of a winning strategy. ©2006 IEEE.