Pub Date : 2019-05-01DOI: 10.1109/iccairo47923.2019.00016
F. Lupiáñez
The aim of this paper is to study some properties of filters in Michálek's fuzzy topological spaces, which are quite different of the classic properties of fuzzy topology. That continues a previous paper of this author.
{"title":"Filters in Michálek's Fuzzy Topological Spaces","authors":"F. Lupiáñez","doi":"10.1109/iccairo47923.2019.00016","DOIUrl":"https://doi.org/10.1109/iccairo47923.2019.00016","url":null,"abstract":"The aim of this paper is to study some properties of filters in Michálek's fuzzy topological spaces, which are quite different of the classic properties of fuzzy topology. That continues a previous paper of this author.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116620506","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00037
George Papageorgiou, G. Demetriou
This paper examines the concepts of learning and diffusion within the context of sustainable mobility and urban development. A System Dynamics (SD) model is proposed, which investigates plausible strategies that can change the mindsets of people towards active mobility. Treating the learning process as a diffusion control process for changing mindsets, is central to the development of the model. Specifically, awareness strategies are investigated as well as introducing Information and Communication Technology (ICT) in a computer simulated environment. The simulation results show that changing mindsets requires time in order to go through the process of knowledge, persuasion, decision, implementation and confirmation which is influenced by the formulated strategies. Carrying out a sensitivity analysis on the various parameters, it was revealed that for changing mindsets effectively, an awareness strategy should be reinforced with the use of an ICT strategy at specific time intervals, during the diffusion process. Optimization on minimizing the required time for developing a sustainable mobility culture is carried out. Such results would be of great use for policy makers interested to promote sustainable mobility such as walking, since their decisions can be tested prior to implementation.
{"title":"Developing a System Dynamics Model for Creating a Learning Sustainable Mobility Culture","authors":"George Papageorgiou, G. Demetriou","doi":"10.1109/ICCAIRO47923.2019.00037","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00037","url":null,"abstract":"This paper examines the concepts of learning and diffusion within the context of sustainable mobility and urban development. A System Dynamics (SD) model is proposed, which investigates plausible strategies that can change the mindsets of people towards active mobility. Treating the learning process as a diffusion control process for changing mindsets, is central to the development of the model. Specifically, awareness strategies are investigated as well as introducing Information and Communication Technology (ICT) in a computer simulated environment. The simulation results show that changing mindsets requires time in order to go through the process of knowledge, persuasion, decision, implementation and confirmation which is influenced by the formulated strategies. Carrying out a sensitivity analysis on the various parameters, it was revealed that for changing mindsets effectively, an awareness strategy should be reinforced with the use of an ICT strategy at specific time intervals, during the diffusion process. Optimization on minimizing the required time for developing a sustainable mobility culture is carried out. Such results would be of great use for policy makers interested to promote sustainable mobility such as walking, since their decisions can be tested prior to implementation.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114553267","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00034
Artemis Chaleplioglou, S. Papavlasopoulos, M. Poulos
BioPortal, the open repository of biomedical ontologies, represents one of the most popular portals for both researchers and practitioners in the Linked Data environment. The BioPortal ontologies contain concepts, relationships, rules and functions to infer the knowledge from various data resources. Solutions of complex biomedical queries is based on the interplay between three types of ontologies: (i) clinical, modelled by SNOMED CT, (ii) pharmacological, modelled by RxNORM, and (iii) genetic, modelled by GO. To explore the degree of integration of BioPortal Ontologies with SNOMED CT, RxNORM and GO ontologies, we collected the BioPortal links and analyzed their connections by descriptive statistics, graphical analysis and agglomerative hierarchical clustering. Whilst nearly all the BioPortal ontologies share links with SNOMED CT, only a quarter out of total share links with RxNORM and only a third out of total share links with GO. A fraction of 3.5% of BioPortal ontologies share links with both RxNORM and GO. Cluster analysis revealed the pattern of ontologies relationships with respect to their links to the SNOMED CT, RxNORM and GO triptych. The NIH, cell biology, pharmacology and chemistry, medical diagnostic and procedure, as well as bibliographic ontologies are clustering together into different subgroups. Collectively, our data suggest, the need for development or enrichment of ontologies connecting all three SNOMED CT, RxNORM and GO. We proposed the usefulness of cluster analysis of linked data to facilitate the selection of closely related ontologies for reuse by the developers.
{"title":"BioPortal Ontologies Integration with SNOMED CT, RxNORM & GO Datasets","authors":"Artemis Chaleplioglou, S. Papavlasopoulos, M. Poulos","doi":"10.1109/ICCAIRO47923.2019.00034","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00034","url":null,"abstract":"BioPortal, the open repository of biomedical ontologies, represents one of the most popular portals for both researchers and practitioners in the Linked Data environment. The BioPortal ontologies contain concepts, relationships, rules and functions to infer the knowledge from various data resources. Solutions of complex biomedical queries is based on the interplay between three types of ontologies: (i) clinical, modelled by SNOMED CT, (ii) pharmacological, modelled by RxNORM, and (iii) genetic, modelled by GO. To explore the degree of integration of BioPortal Ontologies with SNOMED CT, RxNORM and GO ontologies, we collected the BioPortal links and analyzed their connections by descriptive statistics, graphical analysis and agglomerative hierarchical clustering. Whilst nearly all the BioPortal ontologies share links with SNOMED CT, only a quarter out of total share links with RxNORM and only a third out of total share links with GO. A fraction of 3.5% of BioPortal ontologies share links with both RxNORM and GO. Cluster analysis revealed the pattern of ontologies relationships with respect to their links to the SNOMED CT, RxNORM and GO triptych. The NIH, cell biology, pharmacology and chemistry, medical diagnostic and procedure, as well as bibliographic ontologies are clustering together into different subgroups. Collectively, our data suggest, the need for development or enrichment of ontologies connecting all three SNOMED CT, RxNORM and GO. We proposed the usefulness of cluster analysis of linked data to facilitate the selection of closely related ontologies for reuse by the developers.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134406609","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00030
A. Kyriakos, E. Papatheofanous, Bezaitis Charalampos, Evangelos Petrongonas, D. Soudris, D. Reisis
Evolving Convolutional Neural Networks (CNNs) and their execution time performance are key factors for a wide range of applications that are based on deep learning. The need for meeting the applications' time constraints led to design AI accelerators and the current work contributes to this effort by presenting CNN accelerators based on two different design approaches: a) developing CNNs on a power efficient System on Chip (SoC), the Myriad2 and b) a VHDL application specific design and the corresponding FPGA architecture. Both systems target the optimization of time performance regarding the MNIST dataset application. The paper describes the two systems and compares the performance results.
{"title":"Design and Performance Comparison of CNN Accelerators Based on the Intel Movidius Myriad2 SoC and FPGA Embedded Prototype","authors":"A. Kyriakos, E. Papatheofanous, Bezaitis Charalampos, Evangelos Petrongonas, D. Soudris, D. Reisis","doi":"10.1109/ICCAIRO47923.2019.00030","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00030","url":null,"abstract":"Evolving Convolutional Neural Networks (CNNs) and their execution time performance are key factors for a wide range of applications that are based on deep learning. The need for meeting the applications' time constraints led to design AI accelerators and the current work contributes to this effort by presenting CNN accelerators based on two different design approaches: a) developing CNNs on a power efficient System on Chip (SoC), the Myriad2 and b) a VHDL application specific design and the corresponding FPGA architecture. Both systems target the optimization of time performance regarding the MNIST dataset application. The paper describes the two systems and compares the performance results.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"249 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133845928","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00039
Tayfun Topuzoglu, G. Köktürk, D. Altun, A. Şendemir, Ozge Andic Cakır, A. Tokuç, Feyzal Ozkaban
Biologically -inspired computing led to the development of a multi-disciplinary approach and implemented to computing technics. It shows us is that slime mold computes the optimal distance between a food source and its body and then adapts its route. What makes slime mold special is that the computing algorithm changes when exposed to various environmental conditions like repellent or attractive stimuli. In addition, slime mold avoids from the repulsive fields and adopts its route accordingly. The transportation networks can be analyzed by using the optimization capability of slime molds, new transportation lines can be detected, and the efficiency of the existing lines can be analyzed. In this paper, the potential exploitation of the approach for Izmir motorway planning is examined. Slime mold is inoculated on scaled Izmir city map, geographical conditions are represented as attractive and repulsive and the minimum distances are calculated.
{"title":"Finding the Shortest Paths in Izmir Map by Using Slime Molds Images","authors":"Tayfun Topuzoglu, G. Köktürk, D. Altun, A. Şendemir, Ozge Andic Cakır, A. Tokuç, Feyzal Ozkaban","doi":"10.1109/ICCAIRO47923.2019.00039","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00039","url":null,"abstract":"Biologically -inspired computing led to the development of a multi-disciplinary approach and implemented to computing technics. It shows us is that slime mold computes the optimal distance between a food source and its body and then adapts its route. What makes slime mold special is that the computing algorithm changes when exposed to various environmental conditions like repellent or attractive stimuli. In addition, slime mold avoids from the repulsive fields and adopts its route accordingly. The transportation networks can be analyzed by using the optimization capability of slime molds, new transportation lines can be detected, and the efficiency of the existing lines can be analyzed. In this paper, the potential exploitation of the approach for Izmir motorway planning is examined. Slime mold is inoculated on scaled Izmir city map, geographical conditions are represented as attractive and repulsive and the minimum distances are calculated.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"298 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116363495","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00035
Marton Szemenyei, Patrik Reizinger
Several paradigms exist in Reinforcement Learning to improve the exploration capabilities of agents, among which the curiosity-driven approach is followed in this work. Extending previous work that utilizes attention to make curiosity state-and action-selective, we expand the range of experiments by introducing two multi-agent environments. The first one is for robot soccer, the second one features a driving scenario in urban settings. Moreover, as during training the different number of observations must be matched between multiple time-steps, we propose an attention-based approach, called Recurrent Temporal Attention (RTA) to do this. The corresponding implementation can be found at https://github.com/szemenyeim/DynEnv.
{"title":"Attention-Based Curiosity in Multi-Agent Reinforcement Learning Environments","authors":"Marton Szemenyei, Patrik Reizinger","doi":"10.1109/ICCAIRO47923.2019.00035","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00035","url":null,"abstract":"Several paradigms exist in Reinforcement Learning to improve the exploration capabilities of agents, among which the curiosity-driven approach is followed in this work. Extending previous work that utilizes attention to make curiosity state-and action-selective, we expand the range of experiments by introducing two multi-agent environments. The first one is for robot soccer, the second one features a driving scenario in urban settings. Moreover, as during training the different number of observations must be matched between multiple time-steps, we propose an attention-based approach, called Recurrent Temporal Attention (RTA) to do this. The corresponding implementation can be found at https://github.com/szemenyeim/DynEnv.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116404828","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00027
S. Pappas
Climate change and the increased level of power demand has led to a growing electrical energy production from renewable sources, such as wind power. The main problem associated with wind power production is the nature of the wind speed which is random and non linear. This is a reason why wind speed forecasting is a difficult but crucial task, since its accuracy plays a significant role in achieving reliable and autonomous power production and at the same time contributes in surpassing a series problems, of economic and technical nature. In this study real data is used and the performance of three different techniques for adaptive short term wind speed forecasting are evaluated. The first method combines the multimodel partitioning filter (MMPF) implementing extended Kalman filters (EKF) with Support Vector Machines (SVM), the second is a hybrid method of MMPF and Genetic Algorithms (G.A) and the last method implements an artificial multilayer layer feed-forward neural network (ANN). It should be noted that the first two techniques having the structure presented in this work, have never been tested before on short term wind speed prediction. The results indicate that all three methods are reliable, however the combination of MMPF and SVM provides a more accurate wind speed forecasting. Therefore, the proposed method strengthens the prediction precision, and becomes a significant tool for efficient grid planning.
{"title":"Adaptive Forecasting Techniques Applied to Short Time Wind Speed Forecasting","authors":"S. Pappas","doi":"10.1109/ICCAIRO47923.2019.00027","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00027","url":null,"abstract":"Climate change and the increased level of power demand has led to a growing electrical energy production from renewable sources, such as wind power. The main problem associated with wind power production is the nature of the wind speed which is random and non linear. This is a reason why wind speed forecasting is a difficult but crucial task, since its accuracy plays a significant role in achieving reliable and autonomous power production and at the same time contributes in surpassing a series problems, of economic and technical nature. In this study real data is used and the performance of three different techniques for adaptive short term wind speed forecasting are evaluated. The first method combines the multimodel partitioning filter (MMPF) implementing extended Kalman filters (EKF) with Support Vector Machines (SVM), the second is a hybrid method of MMPF and Genetic Algorithms (G.A) and the last method implements an artificial multilayer layer feed-forward neural network (ANN). It should be noted that the first two techniques having the structure presented in this work, have never been tested before on short term wind speed prediction. The results indicate that all three methods are reliable, however the combination of MMPF and SVM provides a more accurate wind speed forecasting. Therefore, the proposed method strengthens the prediction precision, and becomes a significant tool for efficient grid planning.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122114963","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00018
D. Tarasov, O. Milder, A. Tyagunov
Nickel-based superalloys are unique high-temperature materials with complex doping, used, in particular, in gas-turbine engines. These materials exhibit excellent resistance to mechanical and chemical degradation. The main service property of the alloy is its heat resistance, which is expressed, in particular, by the ultimate tensile strength (UTS). When determining the service life of a superalloy product, the developers investigate only certain combinations of temperature parameters and exposure time. The availability of data on the properties of alloys over the entire range of temperatures and time exposures would greatly expand the possibilities of alloys application and would allow more accurate assessment and comparison of alloys. We applied the Bayesian regularized artificial neural network to simulate the missing UTS values for more than 300 well-known superalloys. Network input parameters are the chemical composition and tensile test conditions. Special data pre-processing and a developed learning algorithm significantly reduced the model prediction error. Comparison of the predicted and experimental data showed excellent convergence. A model check was performed on a test data set (10 alloys), which was combined from samples that were not involved in network training.
{"title":"Application of Bayesian Artificial Neural Networks for Modeling the Dependence of Nickel-Based Superalloys' Ultimate Tensile Strength on Their Chemical Composition","authors":"D. Tarasov, O. Milder, A. Tyagunov","doi":"10.1109/ICCAIRO47923.2019.00018","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00018","url":null,"abstract":"Nickel-based superalloys are unique high-temperature materials with complex doping, used, in particular, in gas-turbine engines. These materials exhibit excellent resistance to mechanical and chemical degradation. The main service property of the alloy is its heat resistance, which is expressed, in particular, by the ultimate tensile strength (UTS). When determining the service life of a superalloy product, the developers investigate only certain combinations of temperature parameters and exposure time. The availability of data on the properties of alloys over the entire range of temperatures and time exposures would greatly expand the possibilities of alloys application and would allow more accurate assessment and comparison of alloys. We applied the Bayesian regularized artificial neural network to simulate the missing UTS values for more than 300 well-known superalloys. Network input parameters are the chemical composition and tensile test conditions. Special data pre-processing and a developed learning algorithm significantly reduced the model prediction error. Comparison of the predicted and experimental data showed excellent convergence. A model check was performed on a test data set (10 alloys), which was combined from samples that were not involved in network training.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128299717","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00024
Ivan Ganchev, Zhanlin Ji, M. O'Droma
This paper presents the design aspects of a generic multi-service cloud-based IoT operational platform EMULSION, which is a representative of the new horizontal type, next-generation, IoT platforms that come as a replacement of the existing vertical type platforms. The architectural design and main characteristics of the platform are presented and its multi-tiered structure is explained. Utilizing this EMULSION platform, two pilot platform-demonstration IoT systems are being designed. These along with their performance tests and demonstrations will be reported on in a future paper.
{"title":"A Generic Multi-Service Cloud-Based IoT Operational Platform - EMULSION","authors":"Ivan Ganchev, Zhanlin Ji, M. O'Droma","doi":"10.1109/ICCAIRO47923.2019.00024","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00024","url":null,"abstract":"This paper presents the design aspects of a generic multi-service cloud-based IoT operational platform EMULSION, which is a representative of the new horizontal type, next-generation, IoT platforms that come as a replacement of the existing vertical type platforms. The architectural design and main characteristics of the platform are presented and its multi-tiered structure is explained. Utilizing this EMULSION platform, two pilot platform-demonstration IoT systems are being designed. These along with their performance tests and demonstrations will be reported on in a future paper.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133471062","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 : 2019-05-01DOI: 10.1109/ICCAIRO47923.2019.00023
Marko Ruman, M. Kárný
Finite mixtures of probability densities with components from exponential family serve as flexible parametric models of high-dimensional systems. However, with a few specialized exceptions, these dynamic models assume data-independent weights of mixture components. Their use is illogical and restricts the modeling applicability. The requirement for closeness with respect to conditioning, the basic learning operation, leads to a novel class of models: the mixture ratios. The paper justified them and shows their ability to model truly dynamic systems.
{"title":"Dynamic Mixture Ratio Model","authors":"Marko Ruman, M. Kárný","doi":"10.1109/ICCAIRO47923.2019.00023","DOIUrl":"https://doi.org/10.1109/ICCAIRO47923.2019.00023","url":null,"abstract":"Finite mixtures of probability densities with components from exponential family serve as flexible parametric models of high-dimensional systems. However, with a few specialized exceptions, these dynamic models assume data-independent weights of mixture components. Their use is illogical and restricts the modeling applicability. The requirement for closeness with respect to conditioning, the basic learning operation, leads to a novel class of models: the mixture ratios. The paper justified them and shows their ability to model truly dynamic systems.","PeriodicalId":297342,"journal":{"name":"2019 International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114225780","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：