International Conference on ntelligent Systems Design and Applications, Oct 8, 2015
The design of industrial plant layouts is a complex problem. In order to solve this problem, many... more The design of industrial plant layouts is a complex problem. In order to solve this problem, many authors have used Genetics Algorithms (GAs) with the goal of reaching an efficient plant layout design. The GAs require to define a coding scheme to represent the plant layout design as a data structure. This encoding defines the types of design that can be obtained and it influences heavily in GA ability to find good solutions. This work shows a review of encoding schemes reported in the specialized literature, identify their main characteristics classifying the general types of encoding, and propose approaches that can be applied into the encoding schemesEl diseño de la distribución de áreas en plantas industriales es un problema complejo. Para resolverlo, muchos autores han usado Algoritmos Genéticos (AGs) con el objetivo principal de alcanzar una diseño eficiente de la planta. Los AGs requieren definir un esquema de codificación que represente la distribución de la planta como una estructura de datos. Esta codificación define los tipos de distribución que se pueden obtener y en su fuerte influencia en la capacidad del AG para encontrar buenas soluciones. Este trabajo tiene los siguientes objetivos: presentar una revisión de los esquemas de codificación usados en la literatura, identificar sus principales característica
Revista de Innovación y Buenas Prácticas Docentes, 2018
Los Grados Universitarios en sus planes de estudios tienen asignados una serie de créditos optati... more Los Grados Universitarios en sus planes de estudios tienen asignados una serie de créditos optativos, en los que el estudiante tiene libertad para elegir las asignaturas que más le interesen. Esta elección suele ser bastante complicada para los estudiantes, que si bien cuentan con una guía docente de cada asignatura, que les permite conocer los contenidos teóricos y prácticos, los profesores que la imparten, la metodología y los criterios de evaluación, no suele ser suficiente para decidirse. En este contexto, los estudiantes están interesados en las valoraciones que le pueden ofrecer otros compañeros o compañeras que las cursaron, siendo una práctica habitual que los estudiantes recurran a los estudiantes de otros cursos para que les orienten en el proceso. No obstante, se debe tener cierta precaución con esta práctica, porque no todas las valoraciones que se reciban deberían tener la misma consideración. Por un lado, los intereses y propósitos del estudiante que pide opinión y los...
The use of feature selection can improve accuracy, efficiency, applicability and understandabilit... more The use of feature selection can improve accuracy, efficiency, applicability and understandability of a learning process. For this reason, many methods of automatic feature selection have been developed. Some of these methods are based on the search of the features that allows the data set to be considered consistent. In a search problem we usually evaluate the search states, in the case of feature selection we measure the possible feature sets. This paper reviews the state of the art of consistency based feature selection methods, identifying the measures used for feature sets. An in-deep study of these measures is conducted, including the definition of a new measure necessary for completeness. After that, we perform an empirical evaluation of the measures comparing them with the highly reputed wrapper approach. Consistency measures achieve similar results to those of the wrapper approach with much better efficiency.
This paper deals with the problem of feature selection. Almuallim and Dieterich [1] developed the... more This paper deals with the problem of feature selection. Almuallim and Dieterich [1] developed the FOCUS algorithm which performs optimal feature selection on boolean domains. In a previous paper an extension of FOCUS is developed to deal with discrete and continuous features. In this paper we present an extension to work with fuzzy features, which is verified on two well known problem with quite good results. 1
Feature selection is a mechanism used in Machine Learning to reduce the complexity and improve th... more Feature selection is a mechanism used in Machine Learning to reduce the complexity and improve the speed of the learning process by using a subset of features from the data set. There are several measures which are used to assign a score to a subset of features and, therefore, are able to compare them and decide which one is the best. The bottle neck of consistence measures is having the information of the different examples available to check their class by groups. To handle it, this paper proposes the concept of an algorithmic cache, which stores sorted tables to speed up the access to example information. The work carries out an empirical study using 34 real-world data sets and four representative search strategies combined with different table caching strategies and three sorting methods. The experiments calculate four different consistency and one information measures, showing that the proposed sorted tables cache reduces computation time and it is competitive with hash table structures.
ABSTRACT Introducing expert knowledge into evolutionary algorithms for the facility layout design... more ABSTRACT Introducing expert knowledge into evolutionary algorithms for the facility layout design problem can provide better solutions than the mathematically optimal solutions by considering qualitative aspects in the design. However, this approach requires the direct intervention of a designer (normally called the decision maker) in the evolutionary algorithm that guides the search process to adjust it to his/her preferences. To do this, the designer scores each of the most representative designs of the population to avoid fatigue. The selection of the solutions to be presented for human assessment is crucial, so a small number of solutions that represents the characteristics of the population must be selected without losing the variability of the solutions. The novel hybrid system proposed in this study consists of an interactive genetic algorithm that is combined with two different niching methods to allow interactions between the algorithm and the expert designer. The inclusion of niching techniques into the approach allows for the preservation of diversity, which avoids presenting similar solutions to the designer in the same iteration of the algorithm. The proposed approach was tested using two case studies of facility layout designs. The results of the experiments, which successfully validate the approach, are presented, compared and discussed.
International Conference on ntelligent Systems Design and Applications, Oct 8, 2015
The design of industrial plant layouts is a complex problem. In order to solve this problem, many... more The design of industrial plant layouts is a complex problem. In order to solve this problem, many authors have used Genetics Algorithms (GAs) with the goal of reaching an efficient plant layout design. The GAs require to define a coding scheme to represent the plant layout design as a data structure. This encoding defines the types of design that can be obtained and it influences heavily in GA ability to find good solutions. This work shows a review of encoding schemes reported in the specialized literature, identify their main characteristics classifying the general types of encoding, and propose approaches that can be applied into the encoding schemesEl diseño de la distribución de áreas en plantas industriales es un problema complejo. Para resolverlo, muchos autores han usado Algoritmos Genéticos (AGs) con el objetivo principal de alcanzar una diseño eficiente de la planta. Los AGs requieren definir un esquema de codificación que represente la distribución de la planta como una estructura de datos. Esta codificación define los tipos de distribución que se pueden obtener y en su fuerte influencia en la capacidad del AG para encontrar buenas soluciones. Este trabajo tiene los siguientes objetivos: presentar una revisión de los esquemas de codificación usados en la literatura, identificar sus principales característica
Revista de Innovación y Buenas Prácticas Docentes, 2018
Los Grados Universitarios en sus planes de estudios tienen asignados una serie de créditos optati... more Los Grados Universitarios en sus planes de estudios tienen asignados una serie de créditos optativos, en los que el estudiante tiene libertad para elegir las asignaturas que más le interesen. Esta elección suele ser bastante complicada para los estudiantes, que si bien cuentan con una guía docente de cada asignatura, que les permite conocer los contenidos teóricos y prácticos, los profesores que la imparten, la metodología y los criterios de evaluación, no suele ser suficiente para decidirse. En este contexto, los estudiantes están interesados en las valoraciones que le pueden ofrecer otros compañeros o compañeras que las cursaron, siendo una práctica habitual que los estudiantes recurran a los estudiantes de otros cursos para que les orienten en el proceso. No obstante, se debe tener cierta precaución con esta práctica, porque no todas las valoraciones que se reciban deberían tener la misma consideración. Por un lado, los intereses y propósitos del estudiante que pide opinión y los...
The use of feature selection can improve accuracy, efficiency, applicability and understandabilit... more The use of feature selection can improve accuracy, efficiency, applicability and understandability of a learning process. For this reason, many methods of automatic feature selection have been developed. Some of these methods are based on the search of the features that allows the data set to be considered consistent. In a search problem we usually evaluate the search states, in the case of feature selection we measure the possible feature sets. This paper reviews the state of the art of consistency based feature selection methods, identifying the measures used for feature sets. An in-deep study of these measures is conducted, including the definition of a new measure necessary for completeness. After that, we perform an empirical evaluation of the measures comparing them with the highly reputed wrapper approach. Consistency measures achieve similar results to those of the wrapper approach with much better efficiency.
This paper deals with the problem of feature selection. Almuallim and Dieterich [1] developed the... more This paper deals with the problem of feature selection. Almuallim and Dieterich [1] developed the FOCUS algorithm which performs optimal feature selection on boolean domains. In a previous paper an extension of FOCUS is developed to deal with discrete and continuous features. In this paper we present an extension to work with fuzzy features, which is verified on two well known problem with quite good results. 1
Feature selection is a mechanism used in Machine Learning to reduce the complexity and improve th... more Feature selection is a mechanism used in Machine Learning to reduce the complexity and improve the speed of the learning process by using a subset of features from the data set. There are several measures which are used to assign a score to a subset of features and, therefore, are able to compare them and decide which one is the best. The bottle neck of consistence measures is having the information of the different examples available to check their class by groups. To handle it, this paper proposes the concept of an algorithmic cache, which stores sorted tables to speed up the access to example information. The work carries out an empirical study using 34 real-world data sets and four representative search strategies combined with different table caching strategies and three sorting methods. The experiments calculate four different consistency and one information measures, showing that the proposed sorted tables cache reduces computation time and it is competitive with hash table structures.
ABSTRACT Introducing expert knowledge into evolutionary algorithms for the facility layout design... more ABSTRACT Introducing expert knowledge into evolutionary algorithms for the facility layout design problem can provide better solutions than the mathematically optimal solutions by considering qualitative aspects in the design. However, this approach requires the direct intervention of a designer (normally called the decision maker) in the evolutionary algorithm that guides the search process to adjust it to his/her preferences. To do this, the designer scores each of the most representative designs of the population to avoid fatigue. The selection of the solutions to be presented for human assessment is crucial, so a small number of solutions that represents the characteristics of the population must be selected without losing the variability of the solutions. The novel hybrid system proposed in this study consists of an interactive genetic algorithm that is combined with two different niching methods to allow interactions between the algorithm and the expert designer. The inclusion of niching techniques into the approach allows for the preservation of diversity, which avoids presenting similar solutions to the designer in the same iteration of the algorithm. The proposed approach was tested using two case studies of facility layout designs. The results of the experiments, which successfully validate the approach, are presented, compared and discussed.
Uploads
Papers by Antonio Arauzo-Azofra