Pub Date : 2013-11-01DOI: 10.1109/CINTI.2013.6705209
E. Tóth-Laufer, M. Takács, I. Rudas
In real-time systems the evaluation structure complexity is a key factor, since it depends on the evaluation time whether the result is available in time. In this paper a modification of the conventional Mamdani-type inference system is introduced, which combines the advantages of the Mamdani and Takagi-Sugeno systems. It is computationally cheaper than the original method, thus the response time is shorter, while the intuition can be built into the system. These characteristics make this system suitable to use in real-time risk calculation. This modified inference structure is used in the authors' sport activity risk calculation model. Due to this modification and the use of the HOSVD-based complexity reduction, it can be operated anytime manner to cope with the temporal conditions. In this way the evaluation time can be significantly reduced. The input membership functions are tuned in the system in a patient-specific way to obtain more reliable result. This tuning is performed before the real-time evaluation.
{"title":"Real-time fuzzy logic-based sport activity risk calculation model optimization","authors":"E. Tóth-Laufer, M. Takács, I. Rudas","doi":"10.1109/CINTI.2013.6705209","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705209","url":null,"abstract":"In real-time systems the evaluation structure complexity is a key factor, since it depends on the evaluation time whether the result is available in time. In this paper a modification of the conventional Mamdani-type inference system is introduced, which combines the advantages of the Mamdani and Takagi-Sugeno systems. It is computationally cheaper than the original method, thus the response time is shorter, while the intuition can be built into the system. These characteristics make this system suitable to use in real-time risk calculation. This modified inference structure is used in the authors' sport activity risk calculation model. Due to this modification and the use of the HOSVD-based complexity reduction, it can be operated anytime manner to cope with the temporal conditions. In this way the evaluation time can be significantly reduced. The input membership functions are tuned in the system in a patient-specific way to obtain more reliable result. This tuning is performed before the real-time evaluation.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125112624","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705218
T. Rubina, E. Stalidzans
Biochemical networks of different types such as transcription regulatory, gene regulation, metabolic, protein interaction and signal transduction networks have been extensively studied. One of research directions is related with topological properties of biochemical networks and their structure evolution. Earlier research clearly shows that biochemical networks differ considerably from random ones in their network structure. The reason could be mutation and selection based evolution combined with different importance of particular paths of the network for the survival of an organism. It would be useful to investigate and estimate the changes of network structure that are caused by different types of mutation during the evolution, such as point mutations, missense, nonsense mutations, nucleotide inversion, gene inversion, duplication, deletion, and translocation mutations. BINESA (BIochemical NEtwork Structure Analyser) is a standalone software tool developed for evolution modelling of a biochemical network structure founded on evolutionary changes of an underlying artificial genome. This software tool is written in Visual Basic programming language and developed for Windows operating system using Microsoft Access. BINESA has a database for storing of network structure and artificial genome data. This tool provides the exploration of the evolution dynamics of biochemical network structure, supports SBML and GML models import and export, allows visualization of the network structure in graph form visually marking out reactions of different importance and perform a topological analysis of the biochemical network structure.
{"title":"BINESA — A software tool for evolution modelling of biochemical networks' structure","authors":"T. Rubina, E. Stalidzans","doi":"10.1109/CINTI.2013.6705218","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705218","url":null,"abstract":"Biochemical networks of different types such as transcription regulatory, gene regulation, metabolic, protein interaction and signal transduction networks have been extensively studied. One of research directions is related with topological properties of biochemical networks and their structure evolution. Earlier research clearly shows that biochemical networks differ considerably from random ones in their network structure. The reason could be mutation and selection based evolution combined with different importance of particular paths of the network for the survival of an organism. It would be useful to investigate and estimate the changes of network structure that are caused by different types of mutation during the evolution, such as point mutations, missense, nonsense mutations, nucleotide inversion, gene inversion, duplication, deletion, and translocation mutations. BINESA (BIochemical NEtwork Structure Analyser) is a standalone software tool developed for evolution modelling of a biochemical network structure founded on evolutionary changes of an underlying artificial genome. This software tool is written in Visual Basic programming language and developed for Windows operating system using Microsoft Access. BINESA has a database for storing of network structure and artificial genome data. This tool provides the exploration of the evolution dynamics of biochemical network structure, supports SBML and GML models import and export, allows visualization of the network structure in graph form visually marking out reactions of different importance and perform a topological analysis of the biochemical network structure.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130984865","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705170
Lorinc Martori
This paper addresses the problem of stable bilateral teleoperation in the presence of communication delay by taking into consideration that the outputs of the controllers, which are applied to guarantee the passivity of the teleoperator, are bounded. The velocity and force signals in a teleoperation system are bounded due to the actuator saturation or artificial bounds can also be introduced for the safety and convenience of the human operator. The problem is treated in the time domain passivity approach. It is shown that the passivity of the bilateral teleoperation can be guaranteed with bounded control signals if the force and velocity signals that are sent over the communication channels are also bounded. Real-time experimental measurements were performed to demonstrate the efficiency of the presented method.
{"title":"Time domain passivity based teleoperation applying bounded output controllers","authors":"Lorinc Martori","doi":"10.1109/CINTI.2013.6705170","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705170","url":null,"abstract":"This paper addresses the problem of stable bilateral teleoperation in the presence of communication delay by taking into consideration that the outputs of the controllers, which are applied to guarantee the passivity of the teleoperator, are bounded. The velocity and force signals in a teleoperation system are bounded due to the actuator saturation or artificial bounds can also be introduced for the safety and convenience of the human operator. The problem is treated in the time domain passivity approach. It is shown that the passivity of the bilateral teleoperation can be guaranteed with bounded control signals if the force and velocity signals that are sent over the communication channels are also bounded. Real-time experimental measurements were performed to demonstrate the efficiency of the presented method.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133152621","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705198
A. Stînean, S. Preitl, R. Precup, C. Dragos, M. Radac, Marius-Florin Crainic
Based on our previous research results partially published in [1], [2], [3] and [4], the paper presents a survey on dedicated control solutions for driving systems working under continuously variable conditions: variable reference input (speed), variable moment of inertia and variable load disturbance. The solutions were validated using numerical simulation and tested on a laboratory equipment [5]. The structures employ the switching between different Model-Based (MB) control algorithms; due on the simplicity in adaptation, different fuzzified Takagi-Sugeno control solutions are offered. A hybrid Takagi-Sugeno PI-neuro-fuzzy controller is presented. The solutions are based on a classical cascade control structure with an inner current controller and an external speed control loop with bump-less switching between the control algorithms. Our solutions are representative for mechatronics applications.
基于我们之前在[1]、[2]、[3]、[4]中部分发表的研究成果，本文综述了连续变工况下驱动系统的专用控制方案:变参考输入(速度)、变转动惯量和变负载扰动。通过数值模拟验证了解决方案，并在实验室设备上进行了测试[5]。该结构采用不同的基于模型(MB)控制算法之间的切换;由于适应简单，提出了不同的模糊化Takagi-Sugeno控制方案。提出了一种混合Takagi-Sugeno pi -神经模糊控制器。该方案基于经典的串级控制结构，具有内部电流控制器和外部速度控制环，控制算法之间具有无碰撞切换。我们的解决方案是机电一体化应用的代表。
{"title":"Adaptable fuzzy control solutions for driving systems working under continuously variable conditions","authors":"A. Stînean, S. Preitl, R. Precup, C. Dragos, M. Radac, Marius-Florin Crainic","doi":"10.1109/CINTI.2013.6705198","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705198","url":null,"abstract":"Based on our previous research results partially published in [1], [2], [3] and [4], the paper presents a survey on dedicated control solutions for driving systems working under continuously variable conditions: variable reference input (speed), variable moment of inertia and variable load disturbance. The solutions were validated using numerical simulation and tested on a laboratory equipment [5]. The structures employ the switching between different Model-Based (MB) control algorithms; due on the simplicity in adaptation, different fuzzified Takagi-Sugeno control solutions are offered. A hybrid Takagi-Sugeno PI-neuro-fuzzy controller is presented. The solutions are based on a classical cascade control structure with an inner current controller and an external speed control loop with bump-less switching between the control algorithms. Our solutions are representative for mechatronics applications.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"271 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133712246","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705258
S. Hommel, Darius Malysiak, U. Handmann
In this paper, we describe a method to model human clothes for a later recognition by the use of RGB- and SWIR-cameras. A basic model is estimated during people detection and tracking. This model will be refined if the recognition is triggered. For the refining, several saliency maps are used to extract individual features. These individual features are located separately for any human body parts. The body parts are estimated by the use of a silhouette extraction combined with a skeleton estimation. In this way, the model describes the human clothes in a compact manner which allows the use of a simple and fast comparison method for people recognition. Such models can be used in security and service applications.
{"title":"Model of human clothes based on saliency maps","authors":"S. Hommel, Darius Malysiak, U. Handmann","doi":"10.1109/CINTI.2013.6705258","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705258","url":null,"abstract":"In this paper, we describe a method to model human clothes for a later recognition by the use of RGB- and SWIR-cameras. A basic model is estimated during people detection and tracking. This model will be refined if the recognition is triggered. For the refining, several saliency maps are used to extract individual features. These individual features are located separately for any human body parts. The body parts are estimated by the use of a silhouette extraction combined with a skeleton estimation. In this way, the model describes the human clothes in a compact manner which allows the use of a simple and fast comparison method for people recognition. Such models can be used in security and service applications.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122975016","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705177
Dragan D. Kukolj, D. Dordevic, David Okolišan, I. Ostojic, Dragana D. Sandić-Stanković, C. Hewage
Immersive image/video services will be soon available to the mass market due to the technological advancement of 3D video technologies, which include 3D-Ready TV monitors at affordable prices. However, in order to provide demanding customers with a better service over resource limited (e.g., bandwidth) and unreliable communication channels, system parameters need to be changed “on the fly”. Measured 3D video quality can be used as feedback information to fine tune the system parameters. The main aim of this paper is to analyze and present impact of objective image quality assessment metrics on perception of 3D image/video. Neural Network statistical estimator was used to examine the correlation between objective measures on input image base and Differential Mean Opinion Score (DMOS) of used image base. For this purpose part of LIVE 3D Image Quality Database [7] was used. The results suggest that comparison of the neural network DMOS estimators based on full-reference and no-reference objective metrics shown very similar behavior and accuracy.
{"title":"3D image quality estimation (ANN) based on depth/disparity and 2D metrics","authors":"Dragan D. Kukolj, D. Dordevic, David Okolišan, I. Ostojic, Dragana D. Sandić-Stanković, C. Hewage","doi":"10.1109/CINTI.2013.6705177","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705177","url":null,"abstract":"Immersive image/video services will be soon available to the mass market due to the technological advancement of 3D video technologies, which include 3D-Ready TV monitors at affordable prices. However, in order to provide demanding customers with a better service over resource limited (e.g., bandwidth) and unreliable communication channels, system parameters need to be changed “on the fly”. Measured 3D video quality can be used as feedback information to fine tune the system parameters. The main aim of this paper is to analyze and present impact of objective image quality assessment metrics on perception of 3D image/video. Neural Network statistical estimator was used to examine the correlation between objective measures on input image base and Differential Mean Opinion Score (DMOS) of used image base. For this purpose part of LIVE 3D Image Quality Database [7] was used. The results suggest that comparison of the neural network DMOS estimators based on full-reference and no-reference objective metrics shown very similar behavior and accuracy.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131475528","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705211
Lucian C. Coroianu, R. Fullér
In this paper we study the problem of additivity property of the weighted possibilistic mean operator for interactive fuzzy numbers. We will show that the weighted possibilistic mean operator is additive on the set of symmetric fuzzy numbers if their joint possibility distribution is defined by a t-norm. We will also show some results for general joint-possibility-distribution-based additions of fuzzy numbers of symmetrical opposite sides.
{"title":"On additivity of the weighted possibilistic mean operator","authors":"Lucian C. Coroianu, R. Fullér","doi":"10.1109/CINTI.2013.6705211","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705211","url":null,"abstract":"In this paper we study the problem of additivity property of the weighted possibilistic mean operator for interactive fuzzy numbers. We will show that the weighted possibilistic mean operator is additive on the set of symmetric fuzzy numbers if their joint possibility distribution is defined by a t-norm. We will also show some results for general joint-possibility-distribution-based additions of fuzzy numbers of symmetrical opposite sides.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130391248","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705174
Nasser Lotfi, A. Acan
Multiprocessor scheduling problem is one of the most important issues regarding to parallel programming and distributed system environments. Multiprocessor scheduling is known as a NP-hard problem, hence, applying an exact solution method is not recommended at all. Single-objective type of multiprocessor scheduling problem has already been solved by evolutionary algorithms like genetic algorithms, ant colony optimization, particle swarm optimization, mean field annealing and so on. This paper presents a mean field annealing approach for solving the multi-objective type of this problem. We introduce multi-objective multiprocessor scheduling problem with three objectives and then solve it using mean field annealing approach. Finally, the proposed algorithm is tested over some benchmarks and its effectiveness is compared to NSGA2 and MOGA algorithms. Obtained results show that mean field annealing method leads better Pareto fronts within reasonable computation times.
{"title":"Solving multiprocessor scheduling problem using multi-objective mean field annealing","authors":"Nasser Lotfi, A. Acan","doi":"10.1109/CINTI.2013.6705174","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705174","url":null,"abstract":"Multiprocessor scheduling problem is one of the most important issues regarding to parallel programming and distributed system environments. Multiprocessor scheduling is known as a NP-hard problem, hence, applying an exact solution method is not recommended at all. Single-objective type of multiprocessor scheduling problem has already been solved by evolutionary algorithms like genetic algorithms, ant colony optimization, particle swarm optimization, mean field annealing and so on. This paper presents a mean field annealing approach for solving the multi-objective type of this problem. We introduce multi-objective multiprocessor scheduling problem with three objectives and then solve it using mean field annealing approach. Finally, the proposed algorithm is tested over some benchmarks and its effectiveness is compared to NSGA2 and MOGA algorithms. Obtained results show that mean field annealing method leads better Pareto fronts within reasonable computation times.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"93 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129737848","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705252
Hasan Murat Akinci, E. Yesil
In this study, Fuzzy Cognitive Map (FCM) modeling technique on emotion recognition problem with regression of arousal and valence values is applied and Big Bang - Big Crunch learning method is used for developing the model. Emotions play a critical role of humans' behaviors, beliefs, motivations and decisions. Developing a model between bodily responses and emotional states of a human is an extremely challenging problem in affective computing area. In this study, DEAP dataset, which is publicly available, is used as a dataset. The set contains the recordings of physiological modalities for participant, each participant viewing video clips and reporting emotional states with using self assessment manikins. The results of various simulations show that FCM is a useful and convenient tool for emotion modeling.
本研究将模糊认知图(FCM)建模技术应用于唤醒值和价值回归的情绪识别问题，并采用Big Bang - Big Crunch学习方法开发模型。情绪在人类的行为、信念、动机和决定中起着至关重要的作用。在情感计算领域，建立人的身体反应和情绪状态之间的模型是一个极具挑战性的问题。本研究使用公开的DEAP数据集作为数据集。该集包含了参与者的生理模式记录，每个参与者观看视频剪辑并使用自我评估模型报告情绪状态。各种仿真结果表明，FCM是一种有用且方便的情感建模工具。
{"title":"Emotion modeling using Fuzzy Cognitive Maps","authors":"Hasan Murat Akinci, E. Yesil","doi":"10.1109/CINTI.2013.6705252","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705252","url":null,"abstract":"In this study, Fuzzy Cognitive Map (FCM) modeling technique on emotion recognition problem with regression of arousal and valence values is applied and Big Bang - Big Crunch learning method is used for developing the model. Emotions play a critical role of humans' behaviors, beliefs, motivations and decisions. Developing a model between bodily responses and emotional states of a human is an extremely challenging problem in affective computing area. In this study, DEAP dataset, which is publicly available, is used as a dataset. The set contains the recordings of physiological modalities for participant, each participant viewing video clips and reporting emotional states with using self assessment manikins. The results of various simulations show that FCM is a useful and convenient tool for emotion modeling.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116966239","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 : 2013-11-01DOI: 10.1109/CINTI.2013.6705238
Xiao Yang, Wanlong Chen, Frank Z. Wang
Synaptic plasticity has been widely assumed to be the mechanism behind memory and learning, in which, synapse has a critical role. As a newer biologic update rule to hebbian learning, spiking-time dependent plasticity (STDP) concerns on the temporal order of presynaptic spike and postsynaptic spike which will change the strength of, the connection site of neurons, synapse. In this paper a different way is shown to utilise the novel element memristors to implement a supervised STDP. Because the resistance of memristor depends on its past states, researchers are particularly interested in using such functionality to mimic synaptic connection. Furthermore, benefit from the nano size of memristors and its crossbar structure, large scale neural networks could be implemented. In this supervised STDP, each spike arrival will be assumed to leave a trace which decays exponentially and spikes interact under all-to-all interaction. Depending on the temporal order, memristor synapse will weaken or strengthen the connection of presynaptic neuron and postsynaptic neuron. The temporal all-to-all interaction is implemented during the simulation with training samples. We show that, by combining the memristors, a supervised STDP neural network can be built and learn from the temporal order of presynaptic spike and postsynaptic spike of the training samples.
{"title":"A supervised spiking time dependant plasticity network based on memristors","authors":"Xiao Yang, Wanlong Chen, Frank Z. Wang","doi":"10.1109/CINTI.2013.6705238","DOIUrl":"https://doi.org/10.1109/CINTI.2013.6705238","url":null,"abstract":"Synaptic plasticity has been widely assumed to be the mechanism behind memory and learning, in which, synapse has a critical role. As a newer biologic update rule to hebbian learning, spiking-time dependent plasticity (STDP) concerns on the temporal order of presynaptic spike and postsynaptic spike which will change the strength of, the connection site of neurons, synapse. In this paper a different way is shown to utilise the novel element memristors to implement a supervised STDP. Because the resistance of memristor depends on its past states, researchers are particularly interested in using such functionality to mimic synaptic connection. Furthermore, benefit from the nano size of memristors and its crossbar structure, large scale neural networks could be implemented. In this supervised STDP, each spike arrival will be assumed to leave a trace which decays exponentially and spikes interact under all-to-all interaction. Depending on the temporal order, memristor synapse will weaken or strengthen the connection of presynaptic neuron and postsynaptic neuron. The temporal all-to-all interaction is implemented during the simulation with training samples. We show that, by combining the memristors, a supervised STDP neural network can be built and learn from the temporal order of presynaptic spike and postsynaptic spike of the training samples.","PeriodicalId":439949,"journal":{"name":"2013 IEEE 14th International Symposium on Computational Intelligence and Informatics (CINTI)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126560273","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：