Fitness Function

17:06 – […] a very different kind of approach and a significantly more agile one: instead of defining a specific solution, which I think is what SAFe is trying to do here, in Continuous Delivery we define the Fitness Function, the goals that you are that you should be aiming to iterate towards after each small change. Are you now closer or further from your goal of maintaining your software in a permanently releasable state? That tells you whether the change is a good one or not.

TTITO, Luis Beltran Palma and HUACHO, Javier Arturo Rozas, 2021. Support vector machine for the implementation of the fitness function of genetic algorithms. In: 2021 16th Iberian Conference on Information Systems and Technologies (CISTI). June 2021. p. 1–6. DOI 10.23919/CISTI52073.2021.9476567. A good number of search and optimization problems, computationally speaking, fall within the “intractable problems” and their complexity is of the exponential order. In this type of problem, genetic algorithms are an excellent alternative to find solutions close to the optimum. One of the phases of the genetic algorithms requires a fitness function, which will allow to select the individuals of the next generation. This aptitude function is generally implemented based on mathematical expressions or functions. However, some of the problems that can be solved with genetic algorithms present the difficulty of requiring subjective information to define the fitness function; above all, when the individual selection process requires an assessment conditioned by sensory aspects or experience. In these cases, defining the fitness function mathematically presents a lot of complexity or becomes unfeasible. The present work aims to find computer solutions to this type of problem, embedding the support vector machines in the implementation of the fitness functions of the genetic algorithms.