Multi-population Parallel Genetic Algorithm for Economic Statistical Information Mining Based on Gene Expression Programming

Author(s):  
Qihong Liu ◽  
Tiande Li ◽  
Changjie Tang ◽  
Qiwei Liu ◽  
Chuan Li ◽  
...  
Author(s):  
Baddrud Zaman Laskar ◽  
Swanirbhar Majumder

Gene expression programming (GEP) introduced by Candida Ferreira is a descendant of genetic algorithm (GA) and genetic programming (GP). It takes the advantage of both the optimization and search technique based on genetics and natural selection as GA and its programmatic Darwinian counterpart GP. It is gaining popularity because; it has to some extent eradicated the ‘cons' of both while keeping in the ‘pros'. It is still a new technique not much explored since its introduction in 2001. In this chapter both GA and GP is first discussed followed by the elaborate discussion of GEP. This is followed up by the discussion on research work done is different fields using GEP as a tool followed up by GEP architectures. Finally, here GEP has been used for detection of age from facial features as a soft computing based optimization problem using genetic operators.


Materials ◽  
2019 ◽  
Vol 12 (15) ◽  
pp. 2448 ◽  
Author(s):  
Wang

Concrete producers and construction companies are interested in improving the sustainability of concrete, including reducing its CO2 emissions and the costs of materials while maintaining its mechanical properties, workability, and durability. In this study, we propose a simple approach to the optimal design of the fly ash and slag mixture in blended concrete that considers the carbon pricing, material cost, strength, workability, and carbonation durability. Firstly, the carbon pricing and the material cost are calculated based on the concrete mixture and unit prices. The total cost equals the sum of the material cost and the carbon pricing, and is set as the optimization’s objective function. Secondly, 25 various mixtures are used as a database of optimization. The database covered a wide range of strengths between 25 MPa and 55 MPa and a wide range of workability between 5 and 25 cm in slump. Gene expression programming is used to predict the concrete’s strength and slump. The ternary blended concrete’s carbonation depth is calculated using the efficiency factors of fly ash and slag. Thirdly, the genetic algorithm is used to find the optimal mixture under various constraints. We provide examples to illustrate the design of ternary blended concrete with different strength levels and environmental CO2 concentrations. The results show that, for a suburban region, carbonation durability is the controlling factor in terms of the design of the mixture when the design strength is less than 40.49 MPa, and the compressive strength is the controlling factor in the design of the mixture when the design strength is greater than 40.49 MPa. For an urban region, the critical strength for distinguishing carbonation durability control and strength control is 45.93 MPa. The total cost, material cost, and carbon pricing increase as the concrete’s strength increases.


Author(s):  
Baddrud Zaman Laskar ◽  
Swanirbhar Majumder

Gene expression programming (GEP) introduced by Candida Ferreira is a descendant of genetic algorithm (GA) and genetic programming (GP). It takes the advantage of both the optimization and search technique based on genetics and natural selection as GA and its programmatic Darwinian counterpart GP. It is gaining popularity because; it has to some extent eradicated the ‘cons' of both while keeping in the ‘pros'. It is still a new technique not much explored since its introduction in 2001. In this chapter both GA and GP is first discussed followed by the elaborate discussion of GEP. This is followed up by the discussion on research work done is different fields using GEP as a tool followed up by GEP architectures. Finally, here GEP has been used for detection of age from facial features as a soft computing based optimization problem using genetic operators.


2021 ◽  
Author(s):  
Nigel P. A. Browne

Gene Expression Programming (GEP) is a genetic algorithm that evolves linear chromosomes encoding nonlinear (tree-like) structures. In the original GEP algorithm, the genome size is problem specific and is determined through trial and error. In this work, a novel method for adaptively tuning the genome size is presented. The approach introduces new mutation, transposition and recolI)bination operators that enable a population of heterogeneously structured chromosomes, something the original GEP algorithm does not support. This permits crossbreeding between normally incompatible individuals, speciation within a population, increases the evolvability of the representations and enhances parallel GEP. To test our approach an assortment of problems were used, including symbolic regression, classification and parameter optimization. Our experimental results show that our approach provides a solution for the problem of self-adaptively tuning the genome size of GEP's representation.


ACTA IMEKO ◽  
2012 ◽  
Vol 1 (1) ◽  
pp. 19 ◽  
Author(s):  
Fernando M. Janeiro ◽  
Pedro M. Ramos

Impedance circuit identification through spectroscopy is often used to characterize sensors. When the circuit topology is known, it has been shown that the component values can be obtained by genetic algorithms. Also, gene expression programming can be used to search for an adequate circuit topology. In this paper, an improved version of the impedance circuit identification based on gene expression programming and hybrid genetic algorithm is presented to both identify the circuit and estimate its parameters. Simulation results are used to validate the proposed algorithm in different situations. Further validation is presented from measurements on a circuit that models a humidity sensor and also from measurements on a viscosity sensor.


2013 ◽  
Vol 432 ◽  
pp. 565-570
Author(s):  
Xin Wen Gao ◽  
Ben Bo Guan ◽  
Xing Jian Guan

The purpose of this paper is to improve the efficiency of the Gene Expression Programming (GEP) algorithm. The GEP algorithm is an evolutionary computation. It inherits the characteristics of Genetic Algorithm and Genetic Programming. Through its own characteristics, the GEP algorithm can get the optimal solution of the complicated problem. So, the GEP algorithm has achieved good results in many areas. However, there are also some inevitable drawbacks about the GEP algorithm itself. This paper proposes 5 deficiencies aspects of the GEP algorithm (expression meaning, fitness calculation, local convergence, variable selection, genetic operations, selection of genetic operation rates), and gives the corresponding solutions.


2013 ◽  
Vol 321-324 ◽  
pp. 2223-2226
Author(s):  
Ming Qian Wang ◽  
Wei Tian ◽  
Hao Kang ◽  
Wen Ju Gao

A novel biclustering algorithm is proposed in this paper, which can be used to cluster gene expression data. One of the contributions of this paper is a novel and effective residue function of the biclustering algorithm. Furthermore, a new optimal algorithm which is mixed by the parallel genetic algorithm and the particle swarm optimal algorithm is firstly used to the algorithm of the biclustering for gene expression data. we compared our algorithm with traditional genetic algorithm in biclustering. The results reveal that novel proposed algorithms could discover the interesting patterns in the gene expression profiles.


2021 ◽  
Author(s):  
Nigel P. A. Browne

Gene Expression Programming (GEP) is a genetic algorithm that evolves linear chromosomes encoding nonlinear (tree-like) structures. In the original GEP algorithm, the genome size is problem specific and is determined through trial and error. In this work, a novel method for adaptively tuning the genome size is presented. The approach introduces new mutation, transposition and recolI)bination operators that enable a population of heterogeneously structured chromosomes, something the original GEP algorithm does not support. This permits crossbreeding between normally incompatible individuals, speciation within a population, increases the evolvability of the representations and enhances parallel GEP. To test our approach an assortment of problems were used, including symbolic regression, classification and parameter optimization. Our experimental results show that our approach provides a solution for the problem of self-adaptively tuning the genome size of GEP's representation.


Sign in / Sign up

Export Citation Format

Share Document