Cost friendly Experimental Designs for Product Mixtures in Agricultural Research

Sustainable agriculture practices are the one that caters for the present without compromising for the future generations and at the same time maintains and enhances environmental quality. With the depletion in the quality environmental conditions and an increase in the global population there is an immense need for ensuring sustainability in agricultural practices. Agricultural research can be thought to be as the backbone for bringing out alternative sustainable agricultural methodologies. Precise formulation of agricultural experiments for research requires the application of appropriate statistical tools particularly accurate situation specific designing and analysis of the experiments. There are several experiments in agricultural research where the response depends only on the proportions of the factors in the experiment. The theory of mixture experiments plays a crucial role in such situations and it has wide applicability in agricultural research. One of the difficulties in Mixture Experiments is the generation of design points specific to situation and model chosen. In this study we have attempted to develop an algorithm to obtain designs for mixture experiments specific to the situation. The algorithm is versatile in terms of the situation, number of runs in the design and other parameters and hence has wider applicability. The application of the algorithmic approach of design generation will also lead to minimization in the computational cost involved in the experiments.