Effective texture classification requires image descriptors capable of efficiently detecting, extracting and describing the most relevant information in the images, so that, for instance, different texture classes can be distinguished despite image distortions such as varying illuminations, viewpoints, scales, rotations, etc. Designing such an image descriptor is a challenging task that typically involves the intervention of human experts. In this paper, a general method to automatically design effective image descriptors is proposed. Our method is based on Grammatical Evolution and, using a set of example images from a texture classification problem and a classification algorithm as inputs, generates problem-adapted image descriptors that achieve very competitive classification results. Our method is tested on five well-known texture datasets with different number of classes and image distortions to prove its effectiveness and robustness. Our classification results are statistically compared against those obtained by means of six popular hand-crafted texture descriptors in the state-of-the-art. This statistical analysis shows that our evolutionarily designed descriptors outperform most of those designed by human experts.

Coronavirus disease 2019 (COVID-19) is a highly communicable viral infection caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV2), which has spread rapidly throughout the world. From a computer science point of view, research efforts have focused on the use of approaches such as machine learning and curve fitting to predict or simulate disease behavior. However, the mathematical characterization of the spread of COVID-19 is a topic that has not yet been explored by these techniques. In this work, we propose a novel metaheuristic framework called META-COVID19, which merges the Generalized Boltzmann distribution and the family of Jacobi polynomials to automatically characterize the COVID-19 spread without prior knowledge of the data and without involving a human expert. In general terms, the algorithm receives as input a time series of daily reported cases and the output is a polynomial mathematical model. Our framework only needs a single parameter, which is the number of Jacobi polynomials to analyze during the iterative process, and it is capable of proposing polynomials whose adjustment error is close to 1E-3. Finally, we show the applicability of the polynomial models found by META-COVID19, through a theoretical mathematical analysis in order to know attributes of the spread of COVID-19 in different periods of time, allowing to generate better strategies to face it in the future.