Manageability Challenges for Knowledge Incorporation in Genetic Algorithms: A Study Using the Meal Planning Problem

Abstract

This work investigates manageability challenges for knowledge incorporation in Genetic Algorithms (GAs) for solving the Meal Planning Problem (MPP). The MPP is an intractable problem with optimal solution models in literature leading to solutions that are hard to manage from the knowledge perspective. Manageable incorporation of knowledge into computational models for the MPP will help dietitians in nutrition-based disease therapy administration thereby improving the health of patients. An experimental study implementing a genetic model for the MPP was used to investigate the manageability challenges. The findings were that GAs do not have natural ways of supporting manageable incorporation of knowledge and knowledge incorporated into GAs using existing methods is hard to manage. Manageability is important because knowledge and models become easy to customize to suit different contexts. The novel contribution of this work is a new understanding on the matter of "manageability".