M.Sc. Tezi Görüntüleme | |||||||||||||||||||||
|
|
||||||||||||||||||||
Summary: This work describes the development of an interpreter for the least squares method which is an important technique of regression analysis that fits a mathematical or statistical model to a particular data set, using symbolic computation methods and the JavaCC code generation tool. Although the JavaCC tool is generally used when developing interpreters for programming languages, it can also be used to evaluate mathematical expressions in a similar way. The development process starts with the construction of a context free grammar that denotes the mathematical curves. Then, a parser which is generated via the JavaCC tool for this grammar is employed to represent the curves with object structures and to determine their parameters. Through these object structures, the curves are analyzed and the parameters to be computed by the least squares method are determined. For the curves with specific function components, such as exponential, logarithmic and rational functions, some symbolic computation tasks are performed, which transform those curves into polynomials. Key words: Symbolic computation, Curve fitting, Least squares method, Context-free grammars |