Search

A clustering-based nonlinear ensemble (CNE) learning approach is proposed in this paper to forecast exchange rates. In the proposed CNE learning approach: 1) a self-organizing map neural network is introduced to cluster the in-sample component forecasts; 2) kernel-based extreme learning machine is employed to calculate the in-sample ensemble weights for each cluster; and 3) the corresponding clusters’ in-sample ensemble weights are used for out-of-sample component forecasts to obtain the ensemble forecasts. To illustrate and verify the effectiveness of our proposed model, we test its directional and level forecasting accuracy using four major exchange rates. The out-of-sample forecasting performance results show that the proposed CNE learning approach consistently outperforms the component models and other ensemble learning approaches in terms of the directional forecasting accuracy and the level forecasting accuracy. [ABSTRACT FROM AUTHOR]

Blockchain, as an emerging decentralized architecture and distributed computing paradigm underlying Bitcoin and other cryptocurrencies, has attracted intensive attention in both research and applications recently. Blockchain, especially powered by chain-coded smart contracts, has the full potential of revolutionizing increasingly centralized cyber-physical-social systems (CPSSs) for constructions and applications, and reshaping traditional knowledge automation workflows. The key advantage of blockchain technology lies in the fact that it can enable the establishment of secured, trusted, and decentralized autonomous ecosystems for various scenarios, especially for better usage of the legacy devices, infrastructure, and resources. [ABSTRACT FROM AUTHOR]

Although the introduction of fuzzy theory into a portfolio selection model can help improve the model's practicality, it would increase the difficulty of solving the model. To tackle the issue, this paper proposes a novel mechanism based artificial bee colony algorithm (ABC) consisting of two new proposed learning strategies—direction learning and elite learning. The direction learning strategy has a great potential to guide the search toward the promising areas. The elite learning strategy can gradually pick up the convergence rate without loss of the population diversity. The cooperation of the two approaches forms a mechanism, complementing each other to improve the performance of the algorithms. The proposed mechanism, named LL-mechanism, is introduced into three ABC variants-ABC, gbest-guided ABC (GABC), and CABC, generating LL-ABC, LL-GABC, and LL-CABC, respectively. The experimental results demonstrate the superior performance of the LL-mechanism and LL-CABC outperforms other methods in terms of solution quality, convergence rate, robustness, and numerical stability. Finally, the proposed LL-CABC approach is employed to solve the portfolio selection with fuzzy security return. The experiments on two portfolio selection models illustrate that LL-CABC is effective and promising for a fuzzy portfolio selection. [ABSTRACT FROM AUTHOR]

This paper studies the flexible assembly job-shop scheduling problem in a dynamic manufacturing environment, which is an exension of job-shop scheduling with incorporation of serveral types of flexibilies and integration of an assembly stage. Each product is assembled from several parts with nonlinear process plans with operations involving alternative machines. Setup times are sequence dependent and serparately considered from processing times. Part sharing is fully allowed such that they can be used for the assembly of any possible product, rather than being preassociated to a specfic product. We employ constraint programming and mixed-integer linear programming to formulate the problem. Besides, several dispatching rules with machine feedback machanism are developped. Experimental studies are conducted based on test case problems with different scales and complexities. It is found that constraint programming is the most efficacious approach, whose solution fitness outperforms mixed-integer linear programming as well as all dispatching rules in both static and dynamic cases. On the other hand, dispatching rules are simple to implement, among which the “earliest completion time” rule is the most favourable. A real-time scheduling/rescheduling system has been constructed for the implementation of the proposed approaches to solve practical problems in production. [ABSTRACT FROM AUTHOR]

In this paper, some privacy-preserving features for distributed subgradient optimization algorithms are considered. Most of the existing distributed algorithms focus mainly on the algorithm design and convergence analysis, but not the protection of agents’ privacy. Privacy is becoming an increasingly important issue in applications involving sensitive information. In this paper, we first show that the distributed subgradient synchronous homogeneous-stepsize algorithm is not privacy preserving in the sense that the malicious agent can asymptotically discover other agents’ subgradients by transmitting untrue estimates to its neighbors. Then a distributed subgradient asynchronous heterogeneous-stepsize projection algorithm is proposed and accordingly its convergence and optimality is established. In contrast to the synchronous homogeneous-stepsize algorithm, in the new algorithm agents make their optimization updates asynchronously with heterogeneous stepsizes. The introduced two mechanisms of projection operation and asynchronous heterogeneous-stepsize optimization can guarantee that agents’ privacy can be effectively protected. [ABSTRACT FROM AUTHOR]

This paper studies a mean–variance portfolio selection problem with partial information for a defined-contribution (DC) pension scheme. We assume that the DC pension plan member can only observe the price of the risky asset but not the appreciation rate of it in the financial market. Moreover, inflation risk and salary risk are taken into account in our model. First, by virtue of the filtering theory, we transform the partially observable mean–variance portfolio selection problem to a completely observable one. Then, to look for an equilibrium investment strategy, we formulate and tackle the mean–variance problem within a game theoretic framework. By solving an extended Hamilton–Jacobi–Bellman (HJB) system of equations, closed-form expressions of the equilibrium investment strategy and corresponding equilibrium value function with partial information are derived for the DC pension plan. In addition, some numerical illustrations are provided to show the effects of parameters on the derived equilibrium investment strategies and the efficient frontier. [ABSTRACT FROM PUBLISHER]

Facing horizontal channel competition in a hierarchical distribution system, independent intermediaries such as wholesalers and retailers are keen to find the optimal location and pricing strategies that enable them to adapt to the increasingly competitive business environment. To help market intermediaries to address their challenges, we propose in this paper a spatial agent-based model (SAM), grounded in complex adaptive systems, which comprises four types of agents, namely the world, the manufacturer, firms, and consumers. We derive the firms' optimal behaviors in response to competition by evaluating the evolutionary location and pricing strategies using a genetic algorithm. We observe that a pyramid structure and the bullwhip effect in demand emerge from the evolutionary behavior of the SAM. We also find that buyers' searching ability enhanced by information technology has a significant effect on the degree of competition in a hierarchical distribution system. In addition, we find that firms that distribute elastic goods are likely to lower their prices to attract more buyers and move closer to their suppliers to save transport costs. In the case that the product demand is inelastic, intermediaries will move as close to their buyers as possible because they can maximize their profits in the SAM. [ABSTRACT FROM AUTHOR]

Research project selection is an important task for government and private research funding agencies. When a large number of research proposals are received, it is common to group them according to their similarities in research disciplines. The grouped proposals are then assigned to the appropriate experts for peer review. Current methods for grouping proposals are based on manual matching of similar research discipline areas and/or keywords. However, the exact research discipline areas of the proposals cannot often be accurately designated by the applicants due to their subjective views and possible misinterpretations. Therefore, rich information in the proposals' full text can be used effectively. Text-mining methods have been proposed to solve the problem by automatically classifying text documents, mainly in English. However, these methods have limitations when dealing with non-English language texts, e.g., Chinese research proposals. This paper presents a novel ontology-based text-mining approach to cluster research proposals based on their similarities in research areas. The method is efficient and effective for clustering research proposals with both English and Chinese texts. The method also includes an optimization model that considers applicants' characteristics for balancing proposals by geographical regions. The proposed method is tested and validated based on the selection process at the National Natural Science Foundation of China. The results can also be used to improve the efficiency and effectiveness of research project selection processes in other government and private research funding agencies. [ABSTRACT FROM PUBLISHER]

Motivated by the rapid growth of and an increasing interest in the Chinese online auction market, we empirically investigated EachNet (www.eachnet.com), a highly popular eBay model of a Chinese consumer-to-consumer online auction site. Based on EachNet 's auction mechanisms, a general framework was developed to guide the evaluation of various seller- and buyer-specific factors affecting the outcomes of online auctions, as well as the interrelationships among these factors. Three product-independent performance indicators, namely success-or-failure, effectiveness, and satisfaction, were proposed for assessing the online auction outcome instead of basing assessment on the final price that may vary significantly among different product categories. Our empirical results suggest that the impact of seller characteristics and seller-determined auction attributes on auction performance are partially mediated by the volume of bids generated by all competing bidders during the auction. The implications of this study are also discussed to encourage further research in this area. [ABSTRACT FROM PUBLISHER]