Pub Date : 2018-09-01DOI: 10.1109/SYNASC.2018.00043
Cristina Vatamanu, Dragos Gavrilut, George Popoiu
While machine learning techniques were successfully used for malware identification, they were not without challenges. Over the years, several key points related to the usage of such algorithm for practical applications have evolved: low (close to 0) number of false positives, fast evaluation method, reasonable memory and disk footprint. Because of these constraints, security vendors had to chose a simple algorithm (that can meet all of the above requirements) instead of a more complex ones, even if the later had better detection rates. The present paper describes a hybrid approach that can be used in conjunction with an SVM classifier allowing us to overcome some of the above mentioned constraints.
{"title":"Adjusting SVMs for Large Data Sets using Balanced Decision Trees","authors":"Cristina Vatamanu, Dragos Gavrilut, George Popoiu","doi":"10.1109/SYNASC.2018.00043","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00043","url":null,"abstract":"While machine learning techniques were successfully used for malware identification, they were not without challenges. Over the years, several key points related to the usage of such algorithm for practical applications have evolved: low (close to 0) number of false positives, fast evaluation method, reasonable memory and disk footprint. Because of these constraints, security vendors had to chose a simple algorithm (that can meet all of the above requirements) instead of a more complex ones, even if the later had better detection rates. The present paper describes a hybrid approach that can be used in conjunction with an SVM classifier allowing us to overcome some of the above mentioned constraints.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122722071","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00052
Madalina Erascu, Gabriel Iuhasz, Flavia Micota
Cloud computing offers attractive options to migrate corporate applications without the end users needing to manage any physical resources. While this "ease" is appealing, several issues arise: 1) Which Cloud Providers (CPs) offer the best infrastructure at a fair budget? 2) I am no Cloud expert, then what are the characteristics of the infrastructure which best fit my application? To answer these questions, one must solve a resource management problem, that is, the allocation of computing, storage, networking and (indirectly) cost resources to a set of applications such that the performance objectives of the application are fulfilled. There are many approaches which answer separately these questions but there is no comprehensive and easily usable solution for these issues. MANeUveR solves them by integrating the following components: 1) a Web User Interface offers the end user the possibility to describe his application in terms of interactions between components and their software and hardware requirements, 2) an Offers Management System, through a crawler, periodically updates an ontology with infrastructure and services details from different CPs, 3) a Recommendation Engine provides a (sub) optimal solution for application deployment in the CP infrastructure regarding the leasing price of the virtual machines needed for deployment and their characteristics. Using a secure-billing e-mail service, we demonstrate the effectiveness of our solution.
{"title":"An Architecture for a Management Agency for Cloud Resources","authors":"Madalina Erascu, Gabriel Iuhasz, Flavia Micota","doi":"10.1109/SYNASC.2018.00052","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00052","url":null,"abstract":"Cloud computing offers attractive options to migrate corporate applications without the end users needing to manage any physical resources. While this \"ease\" is appealing, several issues arise: 1) Which Cloud Providers (CPs) offer the best infrastructure at a fair budget? 2) I am no Cloud expert, then what are the characteristics of the infrastructure which best fit my application? To answer these questions, one must solve a resource management problem, that is, the allocation of computing, storage, networking and (indirectly) cost resources to a set of applications such that the performance objectives of the application are fulfilled. There are many approaches which answer separately these questions but there is no comprehensive and easily usable solution for these issues. MANeUveR solves them by integrating the following components: 1) a Web User Interface offers the end user the possibility to describe his application in terms of interactions between components and their software and hardware requirements, 2) an Offers Management System, through a crawler, periodically updates an ontology with infrastructure and services details from different CPs, 3) a Recommendation Engine provides a (sub) optimal solution for application deployment in the CP infrastructure regarding the leasing price of the virtual machines needed for deployment and their characteristics. Using a secure-billing e-mail service, we demonstrate the effectiveness of our solution.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"86 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127653019","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00031
Rodica Condurache
Acacia-K is a tool that solves the synthesis problem for the positive fragment of epistemic temporal logic (KLTL). We briefly describe the implemented algorithm and our test cases. The tool is an extension of Acacia+ that solves the synthesis problem for epistemic temporal specifications where the resulting strategies need memory. To stress more the importance of such implementation, in this paper we compare Acacia-K with MCMAS-SLK, an open-source model-checker supporting the verification of interactive systems against specifications written in a variant of strategy logic under memoryless setting. The results obtained prove the feasibility of our method and represent an encouraging (and necessary) step towards developing implementable procedures for the entire logic KLTL.
Acacia-K是一个解决认知时间逻辑正片段(positive fragment of epistemic temporal logic, KLTL)合成问题的工具。我们简要地描述了实现的算法和我们的测试用例。该工具是Acacia+的扩展，它解决了认知时间规范的综合问题，其中产生的策略需要内存。为了更加强调这种实现的重要性，在本文中，我们将Acacia-K与MCMAS-SLK进行了比较，MCMAS-SLK是一种开源模型检查器，支持根据无内存设置下以策略逻辑变体编写的规范验证交互式系统。所获得的结果证明了我们的方法的可行性，并代表了为整个逻辑KLTL开发可实现程序的一个令人鼓舞的(和必要的)步骤。
{"title":"Experimental Evaluation of Acacia-K: A Tool for Synthesis of Reactive Systems from KLTL+ Specifications","authors":"Rodica Condurache","doi":"10.1109/SYNASC.2018.00031","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00031","url":null,"abstract":"Acacia-K is a tool that solves the synthesis problem for the positive fragment of epistemic temporal logic (KLTL). We briefly describe the implemented algorithm and our test cases. The tool is an extension of Acacia+ that solves the synthesis problem for epistemic temporal specifications where the resulting strategies need memory. To stress more the importance of such implementation, in this paper we compare Acacia-K with MCMAS-SLK, an open-source model-checker supporting the verification of interactive systems against specifications written in a variant of strategy logic under memoryless setting. The results obtained prove the feasibility of our method and represent an encouraging (and necessary) step towards developing implementable procedures for the entire logic KLTL.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124712233","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00020
Kristjan Liiva, Paul B. Jackson, G. Passmore, C. Wintersteiger
We present a compositional validated integration method based on Taylor models. Our method combines solutions for lower dimensional subsystems into solutions for a higher dimensional composite system, rather than attempting to solve the higher dimensional system directly. We have implemented the method in an extension of the Flow* tool. Our preliminary results are promising, suggesting gains for some biological systems with nontrivial compositional structure.
{"title":"Compositional Taylor Model Based Validated Integration","authors":"Kristjan Liiva, Paul B. Jackson, G. Passmore, C. Wintersteiger","doi":"10.1109/SYNASC.2018.00020","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00020","url":null,"abstract":"We present a compositional validated integration method based on Taylor models. Our method combines solutions for lower dimensional subsystems into solutions for a higher dimensional composite system, rather than attempting to solve the higher dimensional system directly. We have implemented the method in an extension of the Flow* tool. Our preliminary results are promising, suggesting gains for some biological systems with nontrivial compositional structure.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125696094","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00069
Alexandru-Ion Marinescu, A. Andreica
Credit scoring is a well established and scrutinized domain within the artificial intelligence field of research and has direct implications in the functioning of financial institutions, by evaluating the risk of approving loans for different clients, which may or may not reimburse them in due time. It is the clients who fail to repay their debt that we are interested in predicting, which makes it a much more difficult task, since they form only a small minority of the total client count. From an input-output perspective, the problem can be stated as: given a set of client properties, such as age, marital status, loan duration, one must yield a 0-1 response variable, with 0 meaning "good" and 1, "bad" clients. Many techniques with high accuracy exist, such as artificial neural networks, but they behave as black box units. We add to this whole context the constraint that the output must be a concrete, tractable mathematical formula, which provides significant added value for a financial analyst. To this end, we present a means for evolving mathematical formulas using genetic programming coupled with Language Integrated Query expression trees, a feature present in the C# programming language.
{"title":"Evolving Mathematical Formulas using LINQ Expression Trees and Direct Applications to Credit Scoring","authors":"Alexandru-Ion Marinescu, A. Andreica","doi":"10.1109/SYNASC.2018.00069","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00069","url":null,"abstract":"Credit scoring is a well established and scrutinized domain within the artificial intelligence field of research and has direct implications in the functioning of financial institutions, by evaluating the risk of approving loans for different clients, which may or may not reimburse them in due time. It is the clients who fail to repay their debt that we are interested in predicting, which makes it a much more difficult task, since they form only a small minority of the total client count. From an input-output perspective, the problem can be stated as: given a set of client properties, such as age, marital status, loan duration, one must yield a 0-1 response variable, with 0 meaning \"good\" and 1, \"bad\" clients. Many techniques with high accuracy exist, such as artificial neural networks, but they behave as black box units. We add to this whole context the constraint that the output must be a concrete, tractable mathematical formula, which provides significant added value for a financial analyst. To this end, we present a means for evolving mathematical formulas using genetic programming coupled with Language Integrated Query expression trees, a feature present in the C# programming language.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134602175","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00021
S. Despréaux, A. Maignan
GPaR is a parallel graph rewriting software implemented in C++ with a graphical user interface. Considering an initial graph g and a system of rewriting rules R = {li->ri, i = 1...n}, GPaR rewrites the graph g into a graph g' by using, simultaneously, the rules of R whose left-hand sides, li, match subgraphs of g. GPaR tackles the problem of overlapping matches and thus can be used in a large variety of rewriting problems including fractal systems. Our proposition is illustrated on the examples of adaptive mesh and Pythagorean tree. The performance of GPaR is compared to the performance of other tools on the Sierpinski triangle benchmark.
GPaR是一个用c++实现的并行图形重写软件，具有图形用户界面。考虑一个初始图g和一个重写规则系统R = {li->ri, i = 1…n}， GPaR通过同时使用R的规则将图g重写为图g'， R的左手边li匹配g的子图。GPaR解决了重叠匹配的问题，因此可以用于包括分形系统在内的各种重写问题。用自适应网格和毕达哥拉斯树的例子说明了我们的命题。将GPaR的性能与Sierpinski三角形基准测试上的其他工具的性能进行比较。
{"title":"GPaR: A Parallel Graph Rewriting Tool","authors":"S. Despréaux, A. Maignan","doi":"10.1109/SYNASC.2018.00021","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00021","url":null,"abstract":"GPaR is a parallel graph rewriting software implemented in C++ with a graphical user interface. Considering an initial graph g and a system of rewriting rules R = {li->ri, i = 1...n}, GPaR rewrites the graph g into a graph g' by using, simultaneously, the rules of R whose left-hand sides, li, match subgraphs of g. GPaR tackles the problem of overlapping matches and thus can be used in a large variety of rewriting problems including fractal systems. Our proposition is illustrated on the examples of adaptive mesh and Pythagorean tree. The performance of GPaR is compared to the performance of other tools on the Sierpinski triangle benchmark.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"26 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130743621","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00045
Ioan Cristian Schuszter
With the emergence of efficient word embeddings for free text, there has been a bloom of Natural Language Processing (NLP) breakthroughs using deep learning techniques. However, the literature is skewed towards the languages that offer large corpuses, and little research has been done in the direction of Romanian text. In this paper, we propose a Deep Learning (DL)-based system for classifying free sentences in the context of psychological surveys, automatically discovering whether respondees are talking about the expected subject in their answers (thoughts, emotions or behaviors) or not. We test several new architectures of convolutional and recurrent neural networks using pre-trained word embeddings from a very large corpus consisting of Wikipedia and Common Crawl data sets. We also present the benefits of transfer learning applied to NLP, allowing general language models trained on large data sets to be applied to problems that have a small amount of data, as in our case, allowing applications of these techniques in many fields.
{"title":"Integrating Deep Learning for NLP in Romanian Psychology","authors":"Ioan Cristian Schuszter","doi":"10.1109/SYNASC.2018.00045","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00045","url":null,"abstract":"With the emergence of efficient word embeddings for free text, there has been a bloom of Natural Language Processing (NLP) breakthroughs using deep learning techniques. However, the literature is skewed towards the languages that offer large corpuses, and little research has been done in the direction of Romanian text. In this paper, we propose a Deep Learning (DL)-based system for classifying free sentences in the context of psychological surveys, automatically discovering whether respondees are talking about the expected subject in their answers (thoughts, emotions or behaviors) or not. We test several new architectures of convolutional and recurrent neural networks using pre-trained word embeddings from a very large corpus consisting of Wikipedia and Common Crawl data sets. We also present the benefits of transfer learning applied to NLP, allowing general language models trained on large data sets to be applied to problems that have a small amount of data, as in our case, allowing applications of these techniques in many fields.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"188 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133772997","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00075
R. Mincu, Alexandru Popa
At CPM 2017, Castelli et al. define and study a new variant of the Longest Common Subsequence Problem, termed the Longest Filled Common Subsequence Problem (LFCS). For the LFCS problem, the input consists of two strings A and B and a multiset of characters M. The goal is to insert the characters from M into the string B, thus obtaining a new string B^*, such that the Longest Common Subsequence (LCS) between A and B^* is maximized. Casteli et al. show that the problem is NP-hard and provide a 3/5-approximation algorithm for the problem. In this paper we study the problem from the experimental point of view. We introduce, implement and test new heuristic algorithms and compare them with the approximation algorithm of Casteli et al. Moreover, we introduce an Integer Linear Program (ILP) model for the problem and we use the state of the art ILP solver, Gurobi, to obtain exact solution for moderate sized instances.
{"title":"Heuristic Algorithms for the Longest Filled Common Subsequence Problem","authors":"R. Mincu, Alexandru Popa","doi":"10.1109/SYNASC.2018.00075","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00075","url":null,"abstract":"At CPM 2017, Castelli et al. define and study a new variant of the Longest Common Subsequence Problem, termed the Longest Filled Common Subsequence Problem (LFCS). For the LFCS problem, the input consists of two strings A and B and a multiset of characters M. The goal is to insert the characters from M into the string B, thus obtaining a new string B^*, such that the Longest Common Subsequence (LCS) between A and B^* is maximized. Casteli et al. show that the problem is NP-hard and provide a 3/5-approximation algorithm for the problem. In this paper we study the problem from the experimental point of view. We introduce, implement and test new heuristic algorithms and compare them with the approximation algorithm of Casteli et al. Moreover, we introduce an Integer Linear Program (ILP) model for the problem and we use the state of the art ILP solver, Gurobi, to obtain exact solution for moderate sized instances.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121030501","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00049
C. Săcărea, Diana Sotropa, Diana Troanca
Formal Concept Analysis (FCA) is a prominent research field having its roots in applied mathematics. Based on a mathematization of concepts and their hierarchies, FCA and its varieties have the potential to unify knowledge discovery methodologies. This paper is devoted to a summary of FCA applications in mining relevant conceptual landscapes from medical data. Electronic Health Records (EHR) constitute a significant technological advance in the way medical information is stored, communicated and processed. Digitized information systems are employed with the aim to improve efficiency, quality of care and costs. We are interested in combining different analysis techniques and visualization methods, such as analogical reasoning, FCA and graph databases in order to bring a fresh perspective over the medical process and to improve the task of knowledge discovery in EHR systems.
{"title":"Formal Concept Analysis Grounded Knowledge Discovery in Electronic Health Record Systems","authors":"C. Săcărea, Diana Sotropa, Diana Troanca","doi":"10.1109/SYNASC.2018.00049","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00049","url":null,"abstract":"Formal Concept Analysis (FCA) is a prominent research field having its roots in applied mathematics. Based on a mathematization of concepts and their hierarchies, FCA and its varieties have the potential to unify knowledge discovery methodologies. This paper is devoted to a summary of FCA applications in mining relevant conceptual landscapes from medical data. Electronic Health Records (EHR) constitute a significant technological advance in the way medical information is stored, communicated and processed. Digitized information systems are employed with the aim to improve efficiency, quality of care and costs. We are interested in combining different analysis techniques and visualization methods, such as analogical reasoning, FCA and graph databases in order to bring a fresh perspective over the medical process and to improve the task of knowledge discovery in EHR systems.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122352195","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 : 2018-09-01DOI: 10.1109/SYNASC.2018.00029
Morad Muslimany, M. Codish
This paper introduces a hybrid search method for optimization problems which combines techniques from Local Search methods and from SAT-based methods. At each iteration, the method performs a "big-step" move on a subset of variables of the current solution. This step is achieved by encoding the big-step itself as an optimization problem and solving it using a SAT (MaxSAT) solver such that the solution of the big-step results in a higher-quality solution to the entire problem. Experimentation illustrates a clear benefit of the approach over both methods: Local Search methods and SAT-based methods.
{"title":"SAT-Based Big-Step Local Search","authors":"Morad Muslimany, M. Codish","doi":"10.1109/SYNASC.2018.00029","DOIUrl":"https://doi.org/10.1109/SYNASC.2018.00029","url":null,"abstract":"This paper introduces a hybrid search method for optimization problems which combines techniques from Local Search methods and from SAT-based methods. At each iteration, the method performs a \"big-step\" move on a subset of variables of the current solution. This step is achieved by encoding the big-step itself as an optimization problem and solving it using a SAT (MaxSAT) solver such that the solution of the big-step results in a higher-quality solution to the entire problem. Experimentation illustrates a clear benefit of the approach over both methods: Local Search methods and SAT-based methods.","PeriodicalId":273805,"journal":{"name":"2018 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)","volume":"178 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122086603","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：