Fitting distribution-like data to exponential sums with genetic algorithms

• Published in: Applied mathematics and computation Vol. 169; no. 1; pp. 82 - 95
• Main Authors: Ma, N.-Y., King, R.P.
• Format: Journal Article
• Language: English
• Published: New York, NY Elsevier Inc 01.10.2005; Elsevier
• Subjects: Algebra; Algorithmics. Computability. Computer arithmetics; Applied sciences; Coding, codes; Computer science; control theory; systems; Distribution-like data; Exact sciences and technology; Exponential-sum functions; Genetic algorithms; Information, signal and communications theory; Mathematics; Number theory; Real-valued coding scheme; Sciences and techniques of general use; Signal and communications theory; Telecommunications and information theory; Theoretical computing; Exponential-sum functions; Real-valued coding scheme; Distribution-like data; Genetic algorithms; Fitting; Exponential sum; Distribution data; Exponential function; Model matching; Genetic algorithm; Coding; Applied mathematics; Trend function
• ISSN: 0096-3003; 1873-5649
• DOI: 10.1016/j.amc.2004.10.036

Conventional derivative-based algorithms of fitting distribution-like data to exponential-sum functions can be easily trapped in some local minima. This paper is concerned with the development of algorithms of fitting distribution-like data to exponential sums with genetic algorithms. Both binary coding scheme and real-valued coding scheme have been investigated in this research. Experimental results have shown that real-valued coding scheme is more appropriate to the problem of fitting distribution-like data to exponential sums. Testing with real engineering data, it has been demonstrated that the fitting algorithm derived in this paper is quite promising. The fitted exponential-sum models using genetic algorithm can very well describe the measured data. However, for the data with wavy trends, pure exponential-sum functions may not be the best candidate models. More generalized exponential-sum models need to be studied.