This research aims to cover a need to be able to classify according to the funds of eyes in diabetic retinopathy disease, how to convert to gray tone, perform an equalization, apply the canny edge highlighting algorithm and apply morphological operations so that a SOM (self-organization map) neural network can be entered and classified. To achieve this, it is classified as 0 to diabetic retinopathy, 1 to glaucoma and 3 to healthy eyes. To corroborate this strategy, a public database of Fundus-images has been taken, being 45 images of eyes for training and for tests 15 images that were not part of the training were used and for the tests 3 images that were not part of the training were used and each grayscale image is scaled to a dimension of 256x256 pixels, managing to demonstrate with this strategy an affectivity of 93.7% certainty in the identification of class of eye disease.

No abstract

Esta investigación tuvo como objetivo encontrar la ruta más corta de n puntos en el espacio, sin corte o intersección entre las líneas generadas por los caminos. Para lograr esto, se usó el algoritmo genético, donde se aprovechó la ventaja de no competir “todos contra todos”, sino que a partir de una pequeña población puede encontrarse la posible mejor ruta en el espacio, también llamada “búsqueda local”. Para realizar el proceso evolutivo, se consideró el método de la ruleta, cruce por intercambio de 2 puntos, mutación por intercambio y método de parada a la varianza a 25 generaciones. Implementado en Matlab 8.3, se obtuvo como resultado una duración de 24.7 s y 210.6 s, con funciones de adaptación de 0.79 y 0.76, 76 y 206 generaciones a las pruebas realizadas de 10 y 100 puntos, respectivamente. Demostrando que con el algoritmo genético se encontró la posible mejor ruta corta de n puntos en el espacio y que el método de parada ideal para este problema en particular es la varianza, aunque consuma más tiempo frente cantidad de generaciones establecidas.

This research aims to cover a need to predict the behavior of the number of deaths in a country that has ended quarantine due to Covid-19.To achieve this, a backpropagation neural network was used as a prediction tool. Taking a public database the country Denmark's death data from Covid-19, with the data accumulated in the range of March 16, 2020 to May 10, 2020 being the inputs for the neural network, managing to predict for May 11, 2020 a cumulative of 209 deaths, implying a forecast of 2 deaths of error according to those that have actually been published of accumulated of 210 deaths. This result represents 98.8% effectiveness.

This article describes a model using genetic algorithm to reorganize the employee in a suitable job position of a company based on their profile (or representation of their resume). In each chromosome the collaborator profile weighting was considered and contrasted with the valuation of the position profile proposed by the representative or head of human resources of the company. Considering that the proposed model is not going to withdraw personnel, nor enter new personnel for the company, just reorganize it. To verify the functionality of the model, the profiles of the 250 collaborators were exchanged with the same 20 profiles proposed by the human resources representative obtaining 0% exchange, when the evaluation was made with their real profiles an exchange was obtained in 14 collaborators, implying 5.6% of collaborators that their profile adapts to another job, delaying for the process 6.2 seconds. This proposed model has a quadratic temporal complexity of 5n2 + 17n + 7.