Welcome to the WIT Press eLibrary - the home of the Transactions of the Wessex Institute collection, providing on-line access to papers presented at the Institute's prestigious international conferences and from its State-of-the-Art in Science & Engineering publications.
There are many methods related to data fitting, and each method has its distinctive features. The article discusses the method of data fitting function under integral criterion. Since the estimate fitting parameters are complicated, the article combines algorithm of simulated annealing and neural network algorithm to solve the integral with neural network algorithm and solve the unknown parameters with simulated annealing algorithm. By case analog computation of household per capita consumption expenditure of urban and the rural residents in China, it proves that the combination of simulated annealing algorithm and neural network algorithm has strong reliability and high accuracy in terms of new method for least absolute integral data fitting.
