Yong-Huang Lin

Dynamic multi-attribute decision making model with grey number evaluations

In practice, multi-period evaluations have to be considered to obtain a reliable decision. Besides, owing to the increasing complexity of decision, the uncertainty of evaluation also increases. Therefore, this study proposes a dynamic decision making model which takes the TOPSIS technique as main structure, integrating the concepts of grey number and Minkowski distance function into it to deal with the uncertain information and aggregate the multi-period evaluations. A subcontractor selection example is adopted to demonstrate the feasibility and practicability of the proposed model. Results show that the proposed model is efficient and robust, and quite good for real-world applications.

Modification of the fish swarm algorithm with particle swarm optimization formulation and communication behavior

The fish swarm algorithm (FSA) is a new intelligent swarm modeling approach that consists primarily of searching, swarming, and following behaviors. This paper proposes several improvements of the FSA, including: (1) using particle swarm optimization formulation to reformulate the FSA, (2) integrating communication behavior into FSA, and (3) creating formulas for major FSA parameters. This paper also focuses on studying the effects of FSA behaviors on optimization during the evolution process. Results focus on the two case study categories of function optimization (eight benchmark functions) and neural network learning (single-input single-output system identification, multi-inputs single output system identification and Iris classification problem). Evidence indicates that the proposed FSA approach reduces the effort necessary to set parameters and that the proposed communication behavior indeed improves FSA.

Adaptive and high-precision grey forecasting model

Although the grey forecasting model has been successfully employed in various fields and demonstrated promising results, literatures show its performance still could be improved. Therefore, a new model named EFGMm(1,1) is proposed in this paper by eliminating the error term resulted from the traditional calculation of background value with an integration equation to substitute for such error term. In addition, Fourier series and exponential smooth technique have also been integrated into the new model to reduce the periodic and stochastic residual errors, respectively. An illustrative example of building material stock index is adopted for demonstration. Results show that the proposed model can increase the prediction accuracy, particularly when the system is instable.

Practical expert diagnosis model based on the grey relational analysis technique

Although the diagnosis problems have been widely discussed in various studies and comprehensively utilized in different fields, literatures show that there still are some limitations in practice. To extend the limitation of application, this study proposes a practical expert diagnosis model which mainly adopts the grey relational analysis technique, a data analytic method based on the generalized distance function, to discriminate the normal objects and abnormal objects. The concept of how the normal objects will be always mapped around a reference point in the multi-dimension space is proposed and explained. Thus the abnormal objects can be identified by the judgement of their distances between the mapped abnormal object and the reference point being exceeded a threshold value. Two verification examples, one is the famous iris data set and the other a slope data set from practical case, are adopted to illustrate the feasibility and applicability of the proposed model in which not only the abnormal objects can be easily distinguished, but also the level of severity of abnormalities can be evaluated.

Gravitational particle swarm

Particle swarm optimization (PSO) is inspired by social behavior of bird flocking, gravitational search algorithm (GSA) is based on the law of gravity, and both of them are related to swarm intelligence (SI). Gravitational particle swarm (GPS) is proposed where a GPS agent has attributes of GSA and PSO. GPS agents update their respective positions with PSO velocity and GSA acceleration. GPS agents, therefore, are able to exhibit PSO bird social and cognitive behaviors and motion in flight, while also reflecting the law of gravity of GSA. From results of 23 benchmark functions, GPS does significantly improve PSO and GSA, with noticeably marked improvements. This paper proposes GPS for hybridizing PSO and GSA due to the outstanding performance and interesting concepts embodied in the GPS.

Applying fuzzy grey modification model on inflow forecasting

This paper investigates a modified grey model for forecasting the inflow of a reservoir. The integral form of the background value is employed for the original grey model, GM(1,1), to improve accuracy and applicability. Thereafter, the Fourier series is altered to handle extreme values with regard to prediction; exponential smoothing is used to improve the drawbacks of the prediction delay phenomenon. Finally, we are hybridised as the ultimate grey model with outstanding prediction accuracy, namely EFGM(1,1). As a typhoon causes significant changes in the inflow of a reservoir, this paper applies the fuzzy membership function for dealing with it during the flood season to construct the fuzzy grey modification model, FEFGM(1,1). Results of grey models are compared with those of the Autoregressive Integrated Moving Average (ARIMA). By evaluating different indices, the errors of the predicted extreme value of EFGM(1,1) perform better than those of GM(1,1) and ARIMA, however worse than that of FEFGM(1,1). The final FEFGM(1,1) shows high precision with regard to reservoir inflow prediction during typhoons with combined effects of fuzzy, exponential smoothing, Fourier series.

Predicting high-strength concrete parameters using weighted genetic programming

Genetic programming (GP) is an evolutionary algorithm-based methodology that employs a binary tree topology with optimized functional operators. This study introduced weight coefficients to each GP linkage in a tree in order to create a new weighted genetic programming (WGP) approach. Two distinct advantages of the proposed WGP include (1) balancing the influences of the two front input branches and (2) incorporating weights throughout generated formulas. Resulting formulas contain a certain quantity of optimized functions and weights. Genetic algorithms are employed to accomplish WGP optimization of function selection and proper weighting tasks. Case studies presented herein highlight a high-strength concrete reference study. Results showed that the proposed WGP not only improves GP in terms of introduced weight coefficients, but also provides both accurate results and formula outputs.

Isolated particle swarm optimization with particle migration and global best adoption

Isolated particle swarm optimization (IPSO) segregates particles into several sub-swarms in order to improve the ability of the global optimization. In this study, particle migration and global best adoption (gbest adoption) are used to improve IPSO. Particle migration allows particles to travel among sub-swarms, based on the fitness of the sub-swarms. The use of gbest adoption allows sub-swarms to peep at the gbest proportionally or probably after a certain number of iterations, i.e. gbest replacing, and gbest sharing, respectively. Three well-known benchmark functions are utilized to determine the parameter settings of the IPSO. Then, 13 benchmark functions are used to study the performance of the designed IPSO. Computational experience demonstrates that the designed IPSO is superior to the original version of particle swarm optimization (PSO) in terms of the accuracy and stability of the results, when isolation phenomenon, particle migration and gbest sharing are involved.

Modular neural network programming with genetic optimization

This study proposes a modular neural network (MNN) that is designed to accomplish both artificial intelligent prediction and programming. Each modular element adopts a high-order neural network to create a formula that considers both weights and exponents. MNN represents practical problems in mathematical terms using modular functions, weight coefficients and exponents. This paper employed genetic algorithms to optimize MNN parameters and designed a target function to avoid over-fitting. Input parameters were identified and modular function influences were addressed in manner that significantly improved previous practices. In order to compare the effectiveness of results, a reference study on high-strength concrete was adopted, which had been previously studied using a genetic programming (GP) approach. In comparison with GP, MNN calculations were more accurate, used more concise programmed formulas, and allowed the potential to conduct parameter studies. The proposed MNN is a valid alternative approach to prediction and programming using artificial neural networks.

Determining ultimate bearing capacity of shallow foundations using a genetic programming system

Three genetic programming models are developed for determining the ultimate bearing capacity of shallow foundations. The proposed genetic programming system (GPS), which comprises genetic programming (GP), weighted genetic programming (WGP), and soft-computing polynomials (SCP), simultaneously provides accurate prediction and visible formulas. Some improvements are achieved for GP and WGP. The SCP is also designed to model the ultimate bearing capacity of shallow foundations with polynomials. Laboratory experimental tests of shallow foundations on cohesionless soils are used with parameters of the angle of shearing resistance, the unit weight of the soil, and the geometry of a foundation considers depth, width, and length to determine the ultimate bearing capacity. Analytical results confirm that all GPS models perform well with acceptable prediction accuracy. Visible formulas of GPS models also facilitate parameter studies, sensitivity analysis, and application of pruning techniques. Notably, SCP gives concise representations for the ultimate bearing capacity and identifies the significant parameters. Although shear resistance angles have the largest impact on ultimate bearing capacity, foundation width and depth are also significant.