The applications of computational algebraic geometry to the analysis of designed experiments: a case study (Q1966366)

This article demonstrates how ideas from algebraic geometry can be introduced to solve certain general problems in the experimental designs for polynomial models. For instance, the present paper discusses following problem: given a particular design deduce a class of models that are estimable with this design. The idea is to replace the natural notion of a design as a set of points in a design space by polynomials whose solutions are the design points. For calculation purposes techniques based on the theory and application of Groebner bases are accessible via computational algebra packages such as Maple and CoCoA. To illustrate the power of the proposed method a real study with non-standard experimental designs is carried out.