Pub Date : 2018-10-01DOI: 10.1109/BRACIS.2018.00025
Rodrigo Exterkoetter, F. Bordignon, L. D. Figueiredo, M. Roisenberg, B. B. Rodrigues
In this paper, we propose a deep convolutional generative adversarial network model to reconstruct the petroleum reservoir connectivity patterns. In the petroleum exploration industry, the critical issue is determining the internal reservoir structure and connectivity, aiming to find a flow channel for placing the injection and the production wells. The state-of-the-art methods propose a combination of seismic inversion with multipoint geostatistics, which imposes connectivity patterns during the optimization. However, this approach has a high computational cost, no learning ability and do not provide a probability through the connection. Results show that our approach is able to learn the petroleum reservoir connectivity patterns from the data and reproduce them also in facies images obtained by the seismic inversion.
{"title":"Petroleum Reservoir Connectivity Patterns Reconstruction Using Deep Convolutional Generative Adversarial Networks","authors":"Rodrigo Exterkoetter, F. Bordignon, L. D. Figueiredo, M. Roisenberg, B. B. Rodrigues","doi":"10.1109/BRACIS.2018.00025","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00025","url":null,"abstract":"In this paper, we propose a deep convolutional generative adversarial network model to reconstruct the petroleum reservoir connectivity patterns. In the petroleum exploration industry, the critical issue is determining the internal reservoir structure and connectivity, aiming to find a flow channel for placing the injection and the production wells. The state-of-the-art methods propose a combination of seismic inversion with multipoint geostatistics, which imposes connectivity patterns during the optimization. However, this approach has a high computational cost, no learning ability and do not provide a probability through the connection. Results show that our approach is able to learn the petroleum reservoir connectivity patterns from the data and reproduce them also in facies images obtained by the seismic inversion.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133990591","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 : 2018-10-01DOI: 10.1109/bracis.2018.00038
Augusto Lopez Dantas, Aurora Trinidad Ramirez Pozo
{"title":"Selecting Algorithms for the Quadratic Assignment Problem with a Multi-label Meta-Learning Approach","authors":"Augusto Lopez Dantas, Aurora Trinidad Ramirez Pozo","doi":"10.1109/bracis.2018.00038","DOIUrl":"https://doi.org/10.1109/bracis.2018.00038","url":null,"abstract":"","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132774419","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00099
Caio Flexa, Reginaldo Santos, W. Gomes, C. Sales
We propose a new non-parametric internal validity index based on mutual equidistant-scattering among within-cluster data for fine-tuning the number of clusters, i.e., the hyperparameter K. Most of the validity indexes found in the literature are considered to be dependent on the number of data objects in clusters and often tend to ignore small and low-density groups. Moreover, they select suboptimal clustering solutions when the clusters are in a certain degree of overlapping or low separation. We analysed our index performance with four of the most popular validity indexes. Experiments on both synthetic and real-world data show the effectiveness and reliability of our approach to evaluate the hyperparameter K.
{"title":"A Novel Equidistant-Scattering-Based Cluster Index","authors":"Caio Flexa, Reginaldo Santos, W. Gomes, C. Sales","doi":"10.1109/BRACIS.2018.00099","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00099","url":null,"abstract":"We propose a new non-parametric internal validity index based on mutual equidistant-scattering among within-cluster data for fine-tuning the number of clusters, i.e., the hyperparameter K. Most of the validity indexes found in the literature are considered to be dependent on the number of data objects in clusters and often tend to ignore small and low-density groups. Moreover, they select suboptimal clustering solutions when the clusters are in a certain degree of overlapping or low separation. We analysed our index performance with four of the most popular validity indexes. Experiments on both synthetic and real-world data show the effectiveness and reliability of our approach to evaluate the hyperparameter K.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130057433","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00079
Bruna Zamith Santos, G. Pereira, F. Nakano, R. Cerri
Transposable Elements (TEs) are DNA sequences capable of changing the gene's activity through transposition within the cells of a host. Once TEs insert themselves in other genes, they can change or reduce the activity of certain proteins, which in some cases could unfeasible the survival of such organisms or even provide genetic variability. A variety of methods has been proposed for the identification and classification of TEs, but most of them still involve a lot of manual work or are too class-specific, which restricts its applicability. Besides, the classes involved in such problems are often hierarchically structured, which is ignored by most of these methods. In this scenario, one problem that still needs further investigation is the use of strategies for selecting positive and negative instances during the induction of hierarchical models. Therefore, in this paper we explore four distinct strategies for selecting training instances, making use of several Machine Learning classifiers with different biases which were applied to the Hierarchical Classification of TEs using a local approach. Thus, we recommend the best strategy based on the results experimentally obtained.
{"title":"Strategies for Selection of Positive and Negative Instances in the Hierarchical Classification of Transposable Elements","authors":"Bruna Zamith Santos, G. Pereira, F. Nakano, R. Cerri","doi":"10.1109/BRACIS.2018.00079","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00079","url":null,"abstract":"Transposable Elements (TEs) are DNA sequences capable of changing the gene's activity through transposition within the cells of a host. Once TEs insert themselves in other genes, they can change or reduce the activity of certain proteins, which in some cases could unfeasible the survival of such organisms or even provide genetic variability. A variety of methods has been proposed for the identification and classification of TEs, but most of them still involve a lot of manual work or are too class-specific, which restricts its applicability. Besides, the classes involved in such problems are often hierarchically structured, which is ignored by most of these methods. In this scenario, one problem that still needs further investigation is the use of strategies for selecting positive and negative instances during the induction of hierarchical models. Therefore, in this paper we explore four distinct strategies for selecting training instances, making use of several Machine Learning classifiers with different biases which were applied to the Hierarchical Classification of TEs using a local approach. Thus, we recommend the best strategy based on the results experimentally obtained.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129145401","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00014
Matheus Carvalho, André Britto
Many-Objective Optimization Problems (MaOPs) are problems that have more than three objective functions to be optimized. Most Multi-Objective Evolutionary Algorithms scales poorly when the number of objective function increases. To face this limitation, new strategies have been proposed. One of them is the use of reference points to enhance the search of the algorithms. NSGA-III is a reference point based algorithm that has been successfully applied to solve MaOPs. NSGA-III uses a set of reference points placed on a normalized hyperplane which is equally inclined to all objective axes and has an intercept at 1 on each axis. Despite the good results of NSGA-III, the shape of the hyper-plane is not deeply explored in literature. This work studies the influence of the set of reference pointsonMany-ObjectiveOptimization.Here, itisproposedthree new transformations of the reference points set used by NSGAIII. Besides, the Vector Guided Adaptation procedure is also applied to modify original NSGA-III hyper-plane. Furthermore, an adaptation of NSGA-III algorithm is proposed and it is performed a set of experiments to evaluate the transformation procedures. Original and adapted versions of NSGA-III are faced over several benchmarking problems observing both convergence and diversity through the analysis of statistical tests.
{"title":"Influence of Reference Points on a Many-Objective Optimization Algorithm","authors":"Matheus Carvalho, André Britto","doi":"10.1109/BRACIS.2018.00014","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00014","url":null,"abstract":"Many-Objective Optimization Problems (MaOPs) are problems that have more than three objective functions to be optimized. Most Multi-Objective Evolutionary Algorithms scales poorly when the number of objective function increases. To face this limitation, new strategies have been proposed. One of them is the use of reference points to enhance the search of the algorithms. NSGA-III is a reference point based algorithm that has been successfully applied to solve MaOPs. NSGA-III uses a set of reference points placed on a normalized hyperplane which is equally inclined to all objective axes and has an intercept at 1 on each axis. Despite the good results of NSGA-III, the shape of the hyper-plane is not deeply explored in literature. This work studies the influence of the set of reference pointsonMany-ObjectiveOptimization.Here, itisproposedthree new transformations of the reference points set used by NSGAIII. Besides, the Vector Guided Adaptation procedure is also applied to modify original NSGA-III hyper-plane. Furthermore, an adaptation of NSGA-III algorithm is proposed and it is performed a set of experiments to evaluate the transformation procedures. Original and adapted versions of NSGA-III are faced over several benchmarking problems observing both convergence and diversity through the analysis of statistical tests.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129851364","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00035
Lucas Borges de Moraes, Adriano Fiorese, R. S. Parpinelli
The cloud computing model has been spreading around the world and has become a basis for innovation and efficiency on provisioning computational services. This fact inspired the emergence of a large number of new companies providing cloud computing services. In order to qualify such providers, performance indicators (PI) are useful for systematic information collection. Select which providers are the most suitable to each customer's needs and with the desired quality of service, has become a hard problem with the need of robust search methods. Thus, the problem is to find the smallest set of providers that maximize the attendance of a customer's request with and the lowest price. In this paper, two evolutionary algorithms, named Genetic Algorithms (GA) and Binary Differential Evolution (BDE), are modeled to address this problem. Instances with 10, 100, and 200 providers are employed. Results obtained are compared with a deterministic method and show that the BDE approach outperforms GA and the deterministic method.
{"title":"Exploring Evolutive Methods for Cloud Provider Selection Based on Performance Indicators","authors":"Lucas Borges de Moraes, Adriano Fiorese, R. S. Parpinelli","doi":"10.1109/BRACIS.2018.00035","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00035","url":null,"abstract":"The cloud computing model has been spreading around the world and has become a basis for innovation and efficiency on provisioning computational services. This fact inspired the emergence of a large number of new companies providing cloud computing services. In order to qualify such providers, performance indicators (PI) are useful for systematic information collection. Select which providers are the most suitable to each customer's needs and with the desired quality of service, has become a hard problem with the need of robust search methods. Thus, the problem is to find the smallest set of providers that maximize the attendance of a customer's request with and the lowest price. In this paper, two evolutionary algorithms, named Genetic Algorithms (GA) and Binary Differential Evolution (BDE), are modeled to address this problem. Instances with 10, 100, and 200 providers are employed. Results obtained are compared with a deterministic method and show that the BDE approach outperforms GA and the deterministic method.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121480756","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00097
Thiago Freitas dos Santos, P. Santos, L. Ferreira, Reinaldo A. C. Bianchi, Pedro Cabalar
Spatial puzzles are interesting domains to investigate problem solving, since the reasoning processes involved in reasoning about spatial knowledge is one of the essential items for an agent to interact in the human environment. With this in mind, the goal of this work is to investigate the knowledge representation and reasoning process related to the solution of a spatial puzzle, the Fisherman's Folly, composed of flexible string, rigid objects and holes. To achieve this goal, the present paper uses heuristics (obtained after solving a relaxed version of the puzzle) to accelerate the learning process, while applying a method that combines Answer Set programming (ASP) with Reinforcement learning (RL), the oASP(MDP) algorithm, to find a solution to the puzzle. ASP is the logic language chosen to build the set of states and actions of a Markov Decision Process (MDP) representing the domain, where RL is used to learn the optimal policy of the problem.
{"title":"Solving a Spatial Puzzle Using Answer Set Programming Integrated with Markov Decision Process","authors":"Thiago Freitas dos Santos, P. Santos, L. Ferreira, Reinaldo A. C. Bianchi, Pedro Cabalar","doi":"10.1109/BRACIS.2018.00097","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00097","url":null,"abstract":"Spatial puzzles are interesting domains to investigate problem solving, since the reasoning processes involved in reasoning about spatial knowledge is one of the essential items for an agent to interact in the human environment. With this in mind, the goal of this work is to investigate the knowledge representation and reasoning process related to the solution of a spatial puzzle, the Fisherman's Folly, composed of flexible string, rigid objects and holes. To achieve this goal, the present paper uses heuristics (obtained after solving a relaxed version of the puzzle) to accelerate the learning process, while applying a method that combines Answer Set programming (ASP) with Reinforcement learning (RL), the oASP(MDP) algorithm, to find a solution to the puzzle. ASP is the logic language chosen to build the set of states and actions of a Markov Decision Process (MDP) representing the domain, where RL is used to learn the optimal policy of the problem.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115877901","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00042
Raphael J. D'Castro, Adriano Oliveira, Augusto H. Terra
Efficiency in the operation is a crucial element for organizations, and this requires greater knowledge about existing business processes. Process Mining is an emerging discipline which aims to provide knowledge about business processes through learning from event logs from the information systems of the organization. Over the last years, the field has progressed significantly, however, some important challenges remain. Among these, we highlight the application in real-world environments, because organizations do not always have well-structured processes, that is, repeatable activities whose inputs and outputs are well defined. Also, there are more complex processes that make it more difficult to get knowledge. This paper presents a study, performed in a real environment, to evaluate the challenges and limitations of the Process Mining tools on process discovery.
{"title":"Process Mining Discovery Techniques in a Low-Structured Process Works?","authors":"Raphael J. D'Castro, Adriano Oliveira, Augusto H. Terra","doi":"10.1109/BRACIS.2018.00042","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00042","url":null,"abstract":"Efficiency in the operation is a crucial element for organizations, and this requires greater knowledge about existing business processes. Process Mining is an emerging discipline which aims to provide knowledge about business processes through learning from event logs from the information systems of the organization. Over the last years, the field has progressed significantly, however, some important challenges remain. Among these, we highlight the application in real-world environments, because organizations do not always have well-structured processes, that is, repeatable activities whose inputs and outputs are well defined. Also, there are more complex processes that make it more difficult to get knowledge. This paper presents a study, performed in a real environment, to evaluate the challenges and limitations of the Process Mining tools on process discovery.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132490161","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00101
Alysson Bispo, R. Prudêncio, D. V. D. Silva
Various software metrics and statistical models have been developed to help companies to predict software defects. Traditional software defect prediction approaches use historical data about previous bugs on a project in order to build predictive machine learning models. However, in many cases the historical testing data available in a project is scarce, i.e., very few or even no labeled training instances are available, which will result on a low quality defect prediction model. In order to overcome this limitation, Cross-Project Defect Prediction (CPDP) can be adopted to learn a defect prediction model for a project of interest (i.e., a target project) by reusing (transferring) data collected from several previous projects (i.e., source projects). In this paper, we focused on neighborhood-based instance selection techniques for CPDP which select labeled instances in the source projects that are similar to the unlabeled instances available in the target project. Despite its simplicity, these techniques have limitations which were addressed in our work. First, although they can select representative source instances, the quality of the selected instances is usually not addressed. Additionally, bug prediction datasets are normally unbalanced (i.e., there are more nondefect instances than defect ones), which can harm learning performance. In this paper, we proposed a new transfer learning approach for CPDP, in which instances selected by a neighborhood-based technique are filtered by the FuzzyRough Instance Selection (FRIS) technique in order to remove noisy instances in the training set. Following, in order to solve class balancing problems, the Synthetic Minority Oversampling Technique (SMOTE) technique is adopted to oversample the minority (defect-prone) class, thus increasing the chance of finding bugs correctly. Experiments were performed on a benchmark set of Java projects, achieving promising results.
{"title":"Instance Selection and Class Balancing Techniques for Cross Project Defect Prediction","authors":"Alysson Bispo, R. Prudêncio, D. V. D. Silva","doi":"10.1109/BRACIS.2018.00101","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00101","url":null,"abstract":"Various software metrics and statistical models have been developed to help companies to predict software defects. Traditional software defect prediction approaches use historical data about previous bugs on a project in order to build predictive machine learning models. However, in many cases the historical testing data available in a project is scarce, i.e., very few or even no labeled training instances are available, which will result on a low quality defect prediction model. In order to overcome this limitation, Cross-Project Defect Prediction (CPDP) can be adopted to learn a defect prediction model for a project of interest (i.e., a target project) by reusing (transferring) data collected from several previous projects (i.e., source projects). In this paper, we focused on neighborhood-based instance selection techniques for CPDP which select labeled instances in the source projects that are similar to the unlabeled instances available in the target project. Despite its simplicity, these techniques have limitations which were addressed in our work. First, although they can select representative source instances, the quality of the selected instances is usually not addressed. Additionally, bug prediction datasets are normally unbalanced (i.e., there are more nondefect instances than defect ones), which can harm learning performance. In this paper, we proposed a new transfer learning approach for CPDP, in which instances selected by a neighborhood-based technique are filtered by the FuzzyRough Instance Selection (FRIS) technique in order to remove noisy instances in the training set. Following, in order to solve class balancing problems, the Synthetic Minority Oversampling Technique (SMOTE) technique is adopted to oversample the minority (defect-prone) class, thus increasing the chance of finding bugs correctly. Experiments were performed on a benchmark set of Java projects, achieving promising results.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126704751","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 : 2018-10-01DOI: 10.1109/BRACIS.2018.00095
T. I. D. Carvalho, Bruno Well Dantas Morais, G. Oliveira
Task scheduling seeks for the time-efficient allocation of the tasks of a parallel program to a multiprocessor system. Being intractable, heuristic methods have been developed to solve this problem. Among the more traditional approaches, approximate techniques as constructive list-based heuristics or simply random schedulers have been extensively employed. On the other hand, bio-inspired models, such as cellular automata (CA) and evolutionary-based schedulers, have been recently investigated as alternative approaches. However, the comparative analysis of the experimental results is primarily limited by the capacity of benchmarks to represent the problem in a full range of difficulty. Aiming to investigate the usage of a more comprehensive benchmark on comparative experiments, we have developed a set of scheduling instances based on real-world programs by applying variations in their features, including number of tasks, number of available processors and communication costs. We have applied two simple heuristics to serve both as baselines for performance and to evaluate the complexity of each problem instance as basis for comparison. Moreover, we investigate here three bio-inspired schedulers applied to the same instances. Two of them are genetic algorithm (GA) approaches while the third employs a GA to find good CA rules able to schedule unseen instances of a parallel program in a very fast operation. Our results show that the CA-based scheduler outperforms the other methods significantly on mosts instances while, on certain instances of the problem, a good solution can be produced consistently by a heuristic based on random allocations. We conclude that these instances are unfit for benchmark purposes and that there is a necessity of careful analysis and selection of problem instances for performance evaluation in this field of research.
{"title":"Bio-Inspired and Heuristic Methods Applied to a Benchmark of the Task Scheduling Problem","authors":"T. I. D. Carvalho, Bruno Well Dantas Morais, G. Oliveira","doi":"10.1109/BRACIS.2018.00095","DOIUrl":"https://doi.org/10.1109/BRACIS.2018.00095","url":null,"abstract":"Task scheduling seeks for the time-efficient allocation of the tasks of a parallel program to a multiprocessor system. Being intractable, heuristic methods have been developed to solve this problem. Among the more traditional approaches, approximate techniques as constructive list-based heuristics or simply random schedulers have been extensively employed. On the other hand, bio-inspired models, such as cellular automata (CA) and evolutionary-based schedulers, have been recently investigated as alternative approaches. However, the comparative analysis of the experimental results is primarily limited by the capacity of benchmarks to represent the problem in a full range of difficulty. Aiming to investigate the usage of a more comprehensive benchmark on comparative experiments, we have developed a set of scheduling instances based on real-world programs by applying variations in their features, including number of tasks, number of available processors and communication costs. We have applied two simple heuristics to serve both as baselines for performance and to evaluate the complexity of each problem instance as basis for comparison. Moreover, we investigate here three bio-inspired schedulers applied to the same instances. Two of them are genetic algorithm (GA) approaches while the third employs a GA to find good CA rules able to schedule unseen instances of a parallel program in a very fast operation. Our results show that the CA-based scheduler outperforms the other methods significantly on mosts instances while, on certain instances of the problem, a good solution can be produced consistently by a heuristic based on random allocations. We conclude that these instances are unfit for benchmark purposes and that there is a necessity of careful analysis and selection of problem instances for performance evaluation in this field of research.","PeriodicalId":405190,"journal":{"name":"2018 7th Brazilian Conference on Intelligent Systems (BRACIS)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114632256","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：