数学建模常用算法（Commonly used mathematical modeling algorithm）.docVIP

数学建模常用算法（Commonly used mathematical modeling algorithm） Methods commonly used in mathematical modeling: analogy method, dichotomy, difference method, variational method, graph theory, analytic hierarchy process, data fitting, regression analysis, mathematical programming (linear programming, nonlinear programming, integer programming, dynamic programming, goal programming), mechanism analysis, queuing method, decision-making methods, countermeasures method, fuzzy evaluation method, time series method, grey theory method and modern optimization algorithms (tabu search algorithm, simulated annealing algorithm, genetic algorithm, neural network). These methods can solve some model: optimization model, differential equation model, statistical model, probabilistic model, graph model, decision model. Fitting and interpolation method (given a number of data points, determine meet the specific requirements of curve or surface, so as to reflect the changing trend of the whole object): Matlab can achieve a function, including polynomial fitting and nonlinear function fitting and multivariate function, namely regression analysis, to determine the function; at the same time can achieve piecewise linear, polynomial spline interpolation, and multidimensional matlab. In the optimization method, decision variables, objective function (simple and smooth), constraint conditions and solving methods are the four key factors. Including non constraint rules (with fminserch, fminbnd (linprog) linear rule), nonlinear rules (implemented in fmincon) multi-objective programming (objective weighting, utility function) dynamic programming, integer programming. Regression analysis: with related phenomenon, according to the relationship between the morphology, the choice of an appropriate mathematical model is used to approximate a statistical method for the average change of the relationship between the variables (the regression linear regression, multiple linear regression, nonlinear regression