Pub Date : 2018-07-08DOI: 10.1109/CEC.2018.8477874
Filipe Assunção, David Sereno, Nuno Lourenço, P. Machado, B. Ribeiro
Developing learning systems is challenging in many ways: often there is the need to optimise the learning algorithm structure and parameters, and it is necessary to decide which is the best data representation to use, i.e., we usually have to design features and select the most representative and useful ones. In this work we focus on the later and investigate whether or not it is possible to obtain good performances with compressed versions of the original data, possibly reducing the learning time. The process of compressing the data, i.e., reducing its dimensionality, is typically conducted by someone who has domain knowledge and expertise, and engineers features in a trial-and-error endless cycle. Our goal is to achieve such compressed versions automatically; for that, we use an Evolutionary Algorithm to generate the structure of AutoEncoders. Instead of targeting the reconstruction of the images, we focus on the reconstruction of the mean signal of each class, and therefore the goal is to acquire the most representative characteristics of each class. Results on the MNIST dataset show that the proposed approach can not only reduce the original dataset dimensionality, but the performance of the classifiers over the compressed representation is superior to the performance on the original uncompressed images.
{"title":"Automatic Evolution of AutoEncoders for Compressed Representations","authors":"Filipe Assunção, David Sereno, Nuno Lourenço, P. Machado, B. Ribeiro","doi":"10.1109/CEC.2018.8477874","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477874","url":null,"abstract":"Developing learning systems is challenging in many ways: often there is the need to optimise the learning algorithm structure and parameters, and it is necessary to decide which is the best data representation to use, i.e., we usually have to design features and select the most representative and useful ones. In this work we focus on the later and investigate whether or not it is possible to obtain good performances with compressed versions of the original data, possibly reducing the learning time. The process of compressing the data, i.e., reducing its dimensionality, is typically conducted by someone who has domain knowledge and expertise, and engineers features in a trial-and-error endless cycle. Our goal is to achieve such compressed versions automatically; for that, we use an Evolutionary Algorithm to generate the structure of AutoEncoders. Instead of targeting the reconstruction of the images, we focus on the reconstruction of the mean signal of each class, and therefore the goal is to acquire the most representative characteristics of each class. Results on the MNIST dataset show that the proposed approach can not only reduce the original dataset dimensionality, but the performance of the classifiers over the compressed representation is superior to the performance on the original uncompressed images.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123991528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-08DOI: 10.1109/CEC.2018.8477900
Karam M. Sallam, S. Elsayed, R. Sarker, D. Essam
Over the last two decades, many different differential evolution (DE) variants have been developed for solving constrained optimization problems. However, none of them performs consistently when solving different types of problems. To deal with this drawback, multiple search operators are used under a single DE algorithm structure where a higher selection pressure is placed on the best performing operator during the evolutionary process. In this paper, we propose to use the landscape information of the problem in the design of the selection mechanism. The performance of this algorithm with the proposed selection mechanism is analysed by solving 10 real-world constrained optimization problems. The experimental results revealed that the proposed algorithm is capable of producing high quality solutions compared to those of state-of-the-art algorithms.
{"title":"Landscape-Based Differential Evolution for Constrained Optimization Problems","authors":"Karam M. Sallam, S. Elsayed, R. Sarker, D. Essam","doi":"10.1109/CEC.2018.8477900","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477900","url":null,"abstract":"Over the last two decades, many different differential evolution (DE) variants have been developed for solving constrained optimization problems. However, none of them performs consistently when solving different types of problems. To deal with this drawback, multiple search operators are used under a single DE algorithm structure where a higher selection pressure is placed on the best performing operator during the evolutionary process. In this paper, we propose to use the landscape information of the problem in the design of the selection mechanism. The performance of this algorithm with the proposed selection mechanism is analysed by solving 10 real-world constrained optimization problems. The experimental results revealed that the proposed algorithm is capable of producing high quality solutions compared to those of state-of-the-art algorithms.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126228892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-08DOI: 10.1109/CEC.2018.8477739
Ali Haidar, B. Verma, Toshi Sinha
Precipitation is viewed as a complex phenomenon that influences the efficiency of the agricultural season. In this paper, an ensemble of neural networks has been created and optimized to estimate monthly rainfall for Innisfail, Australia. The proposed ensemble utilizes single neural networks as components and combines them using a neural network fusion method. A novel ensemble components selection approach has been proposed and deployed. Ensemble components were selected based on a hybrid Genetic Algorithm (GA) that combines standard GA with particle swarm optimization algorithm. Various statistical measurements were calculated to assess the accuracy of the proposed ensembles against single neural networks, climatology and ensembles generated through an alternative selection approach. A better performance was obtained with the proposed ensembles when compared to alternative models.
{"title":"A Novel Approach for Optimizing Ensemble Components in Rainfall Prediction","authors":"Ali Haidar, B. Verma, Toshi Sinha","doi":"10.1109/CEC.2018.8477739","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477739","url":null,"abstract":"Precipitation is viewed as a complex phenomenon that influences the efficiency of the agricultural season. In this paper, an ensemble of neural networks has been created and optimized to estimate monthly rainfall for Innisfail, Australia. The proposed ensemble utilizes single neural networks as components and combines them using a neural network fusion method. A novel ensemble components selection approach has been proposed and deployed. Ensemble components were selected based on a hybrid Genetic Algorithm (GA) that combines standard GA with particle swarm optimization algorithm. Various statistical measurements were calculated to assess the accuracy of the proposed ensembles against single neural networks, climatology and ensembles generated through an alternative selection approach. A better performance was obtained with the proposed ensembles when compared to alternative models.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129954981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477771
Marwa Hammami, Slim Bechikh, C. Hung, L. B. Said
Feature selection and construction are important pre-processing techniques in data mining. They may allow not only dimensionality reduction but also classifier accuracy and efficiency improvement. These two techniques are of great importance especially for the case of high-dimensional data. Feature construction for high-dimensional data is still a very challenging topic. This can be explained by the large search space of feature combinations, whose size is a function of the number of features. Recently, researchers have used Genetic Programming (GP) for feature construction and the obtained results were promising. Unfortunately, the wrapper evaluation of each feature subset, where a feature can be constructed by a combination of features, is computationally intensive since such evaluation requires running the classifier on the data sets. Motivated by this observation, we propose, in this paper, a hybrid multiobjective evolutionary approach for efficient feature construction and selection. Our approach uses two filter objectives and one wrapper objective corresponding to the accuracy. In fact, the whole population is evaluated using two filter objectives. However, only non-dominated (best) feature subsets are improved using an indicator-based local search that optimizes the three objectives simultaneously. Our approach has been assessed on six high-dimensional datasets and compared with two existing prominent GP approaches, using three different classifiers for accuracy evaluation. Based on the obtained results, our approach is shown to provide competitive and better results compared with two competitor GP algorithms tested in this study.
{"title":"A Multi-Objective Hybrid Filter-Wrapper Evolutionary Approach for Feature Construction on High-Dimensional Data","authors":"Marwa Hammami, Slim Bechikh, C. Hung, L. B. Said","doi":"10.1109/CEC.2018.8477771","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477771","url":null,"abstract":"Feature selection and construction are important pre-processing techniques in data mining. They may allow not only dimensionality reduction but also classifier accuracy and efficiency improvement. These two techniques are of great importance especially for the case of high-dimensional data. Feature construction for high-dimensional data is still a very challenging topic. This can be explained by the large search space of feature combinations, whose size is a function of the number of features. Recently, researchers have used Genetic Programming (GP) for feature construction and the obtained results were promising. Unfortunately, the wrapper evaluation of each feature subset, where a feature can be constructed by a combination of features, is computationally intensive since such evaluation requires running the classifier on the data sets. Motivated by this observation, we propose, in this paper, a hybrid multiobjective evolutionary approach for efficient feature construction and selection. Our approach uses two filter objectives and one wrapper objective corresponding to the accuracy. In fact, the whole population is evaluated using two filter objectives. However, only non-dominated (best) feature subsets are improved using an indicator-based local search that optimizes the three objectives simultaneously. Our approach has been assessed on six high-dimensional datasets and compared with two existing prominent GP approaches, using three different classifiers for accuracy evaluation. Based on the obtained results, our approach is shown to provide competitive and better results compared with two competitor GP algorithms tested in this study.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115706694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477919
M. Cauwet, O. Teytaud
This paper studies the optimization of strategies in the context of possibly randomized two players zero-sum games with incomplete information. We compare 5 algorithms for tuning the parameters of strategies over a benchmark of 12 games. A first evolutionary approach consists in designing a highly randomized opponent (called naive opponent) and optimizing the parametric strategy against it; a second one is optimizing iteratively the strategy, i.e. constructing a sequence of strategies starting from the naive one. 2 versions of coevolutions, real and approximate, are also tested as well as a seed method. The coevolution methods were performing well, but results were not stable from one game to another. In spite of its simplicity, the seed method, which can be seen as an extremal version of coevolution, works even when nothing else works. Incidentally, these methods brought out some unexpected strategies for some games, such as Batawaf or the game of War, which seem, at first view, purely random games without any structured actions possible for the players or Guess Who, where a dichotomy between the characters seems to be the most reasonable strategy. All source codes of games are written in Matlab/Octave and are freely available for download.
{"title":"Surprising Strategies Obtained by Stochastic Optimization in Partially Observable Games","authors":"M. Cauwet, O. Teytaud","doi":"10.1109/CEC.2018.8477919","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477919","url":null,"abstract":"This paper studies the optimization of strategies in the context of possibly randomized two players zero-sum games with incomplete information. We compare 5 algorithms for tuning the parameters of strategies over a benchmark of 12 games. A first evolutionary approach consists in designing a highly randomized opponent (called naive opponent) and optimizing the parametric strategy against it; a second one is optimizing iteratively the strategy, i.e. constructing a sequence of strategies starting from the naive one. 2 versions of coevolutions, real and approximate, are also tested as well as a seed method. The coevolution methods were performing well, but results were not stable from one game to another. In spite of its simplicity, the seed method, which can be seen as an extremal version of coevolution, works even when nothing else works. Incidentally, these methods brought out some unexpected strategies for some games, such as Batawaf or the game of War, which seem, at first view, purely random games without any structured actions possible for the players or Guess Who, where a dichotomy between the characters seems to be the most reasonable strategy. All source codes of games are written in Matlab/Octave and are freely available for download.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"261 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123103110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477693
Tobias Rodemann
In industry we see an increasing interest in (evolutionary) many objective optimization algorithms. However, the majority of engineers only using, not researching, optimizers have a limited understanding of the pros and cons of different algorithms and therefore rely on either third-party recommendations or benchmark tests to pick the most suitable methods for their problems. Unfortunately, most benchmarks are targeting an academic audience leaving the practitioner often in doubt about the correct choices. In this article we try to outline the essential requirements for a many-objective optimization algorithm portfolio management from an industrial perspective and compare the situation in our field to another domain with similar issues, image processing. We want to address one of the core practical issues: “Given a limited computational or time budget for my optimization project, which optimization algorithms should I try?”.
{"title":"Industrial Portfolio Management for Many-Objective Optimization Algorithms","authors":"Tobias Rodemann","doi":"10.1109/CEC.2018.8477693","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477693","url":null,"abstract":"In industry we see an increasing interest in (evolutionary) many objective optimization algorithms. However, the majority of engineers only using, not researching, optimizers have a limited understanding of the pros and cons of different algorithms and therefore rely on either third-party recommendations or benchmark tests to pick the most suitable methods for their problems. Unfortunately, most benchmarks are targeting an academic audience leaving the practitioner often in doubt about the correct choices. In this article we try to outline the essential requirements for a many-objective optimization algorithm portfolio management from an industrial perspective and compare the situation in our field to another domain with similar issues, image processing. We want to address one of the core practical issues: “Given a limited computational or time budget for my optimization project, which optimization algorithms should I try?”.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117017276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477861
P. Rakshit, S. Saha, A. Konar, A. Nagar
The proposed system aims at automatic identification of an unknown dance posture referring to the 34 primitive postures of ballet, simultaneously measuring the proximity of an unknown dance posture to a known primitive. A simple and novel seven stage algorithm achieves the desired objective. Skin color segmentation is performed on the dance postures, the output of which is dilated and edge is detected. From the boundaries of the postures, connected components are identified and the boundary is piecewise linearly approximated using modified artificial bee colony algorithm. Here, lies the novelty of our work. From the approximated boundary, features are extracted in terms of internal angles. This whole procedure is repeated for all the training images as well as testing image. The classification of the training image containing ballet posture is done using Euclidean distance matching.
{"title":"Evolutionary Approach to Straight Line Approximation for Image Matching in Dance-Posture Recognition","authors":"P. Rakshit, S. Saha, A. Konar, A. Nagar","doi":"10.1109/CEC.2018.8477861","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477861","url":null,"abstract":"The proposed system aims at automatic identification of an unknown dance posture referring to the 34 primitive postures of ballet, simultaneously measuring the proximity of an unknown dance posture to a known primitive. A simple and novel seven stage algorithm achieves the desired objective. Skin color segmentation is performed on the dance postures, the output of which is dilated and edge is detected. From the boundaries of the postures, connected components are identified and the boundary is piecewise linearly approximated using modified artificial bee colony algorithm. Here, lies the novelty of our work. From the approximated boundary, features are extracted in terms of internal angles. This whole procedure is repeated for all the training images as well as testing image. The classification of the training image containing ballet posture is done using Euclidean distance matching.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127087048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477788
L. A. Soares, K. F. Côco, E. Salles, P. M. Ciarelli
This work proposes a new method of representing textures on digital images through their maximum (or minimum if the negative of the image is used) and different intensity borders by means of artificial beings called artificial hikers that search for the maximum of a texture image and in doing so, represent the different characteristics of the image. The technique has two main parameters that can be adjusted in order to emphasize the greatest maximum of an image and different frequency borders. The results show that it is a very flexible technique on representing different components of a texture image, working on both natural and artificial images. For the classification of textures, the technique of artificial hikers was combined with fractal dimension analysis and it presented superior results compared to previous works dealing with texture classification with artificial agents.
{"title":"Texture Representation and Classification with Artificial Hikers and Fractals","authors":"L. A. Soares, K. F. Côco, E. Salles, P. M. Ciarelli","doi":"10.1109/CEC.2018.8477788","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477788","url":null,"abstract":"This work proposes a new method of representing textures on digital images through their maximum (or minimum if the negative of the image is used) and different intensity borders by means of artificial beings called artificial hikers that search for the maximum of a texture image and in doing so, represent the different characteristics of the image. The technique has two main parameters that can be adjusted in order to emphasize the greatest maximum of an image and different frequency borders. The results show that it is a very flexible technique on representing different components of a texture image, working on both natural and artificial images. For the classification of textures, the technique of artificial hikers was combined with fractal dimension analysis and it presented superior results compared to previous works dealing with texture classification with artificial agents.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125071237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477741
R. Šenkeřík, Adam Viktorin, Michal Pluhacek, T. Kadavy
This research deals with the modern and popular hybridization of chaotic dynamics and evolutionary computation. Unlike many research studies on the combination of chaos and metaheuristics, this paper focuses on the deeper insight into the population dynamics, specifically influence of chaotic sequences on the population diversity. The optimization algorithm performance was recorded as well. Experiments are focused on the extensive investigation of the different randomization schemes for the selection of individuals in a simple parameter adaptive Differential Evolution (DE) strategy: jDE algorithm. The jDE was driven by the nine different two-dimensional discrete chaotic systems, as the chaotic pseudo-random number generators. The population diversity and jDE convergence are recorded on the two-dimensional settings (10D and 30D) and 15 test functions from the CEC 2015 benchmark.
{"title":"On the Population Diversity for the Chaotic Differential Evolution","authors":"R. Šenkeřík, Adam Viktorin, Michal Pluhacek, T. Kadavy","doi":"10.1109/CEC.2018.8477741","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477741","url":null,"abstract":"This research deals with the modern and popular hybridization of chaotic dynamics and evolutionary computation. Unlike many research studies on the combination of chaos and metaheuristics, this paper focuses on the deeper insight into the population dynamics, specifically influence of chaotic sequences on the population diversity. The optimization algorithm performance was recorded as well. Experiments are focused on the extensive investigation of the different randomization schemes for the selection of individuals in a simple parameter adaptive Differential Evolution (DE) strategy: jDE algorithm. The jDE was driven by the nine different two-dimensional discrete chaotic systems, as the chaotic pseudo-random number generators. The population diversity and jDE convergence are recorded on the two-dimensional settings (10D and 30D) and 15 test functions from the CEC 2015 benchmark.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123475814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-07-01DOI: 10.1109/CEC.2018.8477855
T. Lucas, Renato Vimieiro, Teresa B Ludermir
This paper presents an evolutionary approach for mining diverse and more informative subgroups focused on high dimensional data sets. Subgroup Discovery (SD) is an important tool for knowledge discovery that aims to identify sets of features that distinguish a target group from the others (e.g. successful from unsuccessful treatments). At the same time, to extract information from high dimensional data sets becomes more natural. One of the first and most efficient SD heuristics focused on high dimensional data is the SSDP. However, this model deals superficially with diverse/redundancy in top-k subgroups, which can result in poor information for users. This work presents SSDP+, an extension of the SSDP model to provide diversity in a way that explore the relation between subgroups order to 2enerate a more informative set of patterns.
{"title":"SSDP+: A Diverse and More Informative Subgroup Discovery Approach for High Dimensional Data","authors":"T. Lucas, Renato Vimieiro, Teresa B Ludermir","doi":"10.1109/CEC.2018.8477855","DOIUrl":"https://doi.org/10.1109/CEC.2018.8477855","url":null,"abstract":"This paper presents an evolutionary approach for mining diverse and more informative subgroups focused on high dimensional data sets. Subgroup Discovery (SD) is an important tool for knowledge discovery that aims to identify sets of features that distinguish a target group from the others (e.g. successful from unsuccessful treatments). At the same time, to extract information from high dimensional data sets becomes more natural. One of the first and most efficient SD heuristics focused on high dimensional data is the SSDP. However, this model deals superficially with diverse/redundancy in top-k subgroups, which can result in poor information for users. This work presents SSDP+, an extension of the SSDP model to provide diversity in a way that explore the relation between subgroups order to 2enerate a more informative set of patterns.","PeriodicalId":212677,"journal":{"name":"2018 IEEE Congress on Evolutionary Computation (CEC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125357025","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：