Papers by Nasser Mozayani
International Conference on Artificial Intelligence, 2008
In this paper we propose a new gradual noisy chaotic neural network (MP-NCNN) to solve the NP-com... more In this paper we propose a new gradual noisy chaotic neural network (MP-NCNN) to solve the NP-complete attributed relational graph matching problem. These graphs are very important in pattern matching applications and the noisy chaotic behavior of the proposed method which avoids getting trapped in local minima, yields in better results and hence it is more effective approach in comparison with previous methods. The performance of the proposed method has been evaluated through several attributed relational graphs with different permuted vertices and also with different vertex numbers. The obtained results show that the proposed method outperforms previous approaches including HNN, CNN and TCNN methods.
In this paper we propose an algorithm for converting dependency structures to phrase structures. ... more In this paper we propose an algorithm for converting dependency structures to phrase structures. This algorithm mainly concerns the characteristics of non-configurational languages. We review current works in the field and on the basis of these works we try to adopt a more flexible approach to the problem.
International journal of computer applications, Jan 16, 2014
Particle swarm optimization is a population-based algorithm and used for optimization in a wide r... more Particle swarm optimization is a population-based algorithm and used for optimization in a wide range of problems. In this article, a method that is called Hybrid Particle Swarm Optimization or HPSO is proposed. It is composed of some versions of particle swarm optimization algorithms, which have subgroups in their structures. They are DMS-PSO, PS2OS and MCPSO. In fact, a hierarchical structure is used to compose a new version of optimization algorithm and combine the results of other structures of PSO. Proposed structure has been tested on four unimodal and four multimodal test functions. Although the memory usage has no difference with other compared versions, it is much faster in many cases. Also the rank of fitness values, are good and suitable in all test functions. In addition, it is possible to execute it concurrently.
POMCP-based decentralized spatial task allocation algorithms for partially observable environments
Applied Intelligence
Multi-agent Q-Learning control of spacecraft formation flying reconfiguration trajectories
Advances in Space Research
Optical Character Recognition (OCR) is a very old and of great interest in pattern recognition fi... more Optical Character Recognition (OCR) is a very old and of great interest in pattern recognition field. In this paper we introduce a very powerful approach to recognize Persian text. We have used morphological operators, especially Hit/Miss operator to descript each sub-word and by using a template matching approach we have tried to classify generated description. We used just one font in two different sizes to verify our approach. We achieved a very good rate, up to 99.9%.
HOMAN, a learning based negotiation method for holonic multi-agent systems
Journal of Intelligent & Fuzzy Systems, 2014
ABSTRACT Holonic multi-agent systems are a special category of multi-agent systems that best fit ... more ABSTRACT Holonic multi-agent systems are a special category of multi-agent systems that best fit to environments with numerous agents and high complexity. Like in general multi-agent systems, the agents in the holonic system may negotiate with each other. These systems have their own characteristics and structure, for which a specific negotiation mechanism is required. This mechanism should be simple, fast and operable in real world applications. It would be better to equip negotiators with a learning method which can efficiently use the available information. The learning method should itself be fast, too. Additionally, this mechanism should match the special characteristics of the holonic multi-agent systems. In this paper, we introduce such a negotiation method. Experimental results demonstrate the efficiency of this new approach.
Constructing and Evaluating Options in Reinforcement Learning
2018 9th International Symposium on Telecommunications (IST), 2018
In this paper, we propose a new subgoal based method for automatic construction of useful options... more In this paper, we propose a new subgoal based method for automatic construction of useful options. In our proposed method, subgoals are considered as border states of communities of the transition graph created after some initial agent interactions with the environment. We present a new community detection algorithm to provide an appropriate partitioning of the transition graph. Macro-actions are constructed for taking the agent from one community to other communities. In addition, we attempt to capture intuitions about features of useful macro-actions. There is a lack of a generic evaluation mechanism for evaluating each macro-action in previous research. We will propose a method for evaluating each macro-action separately. Inappropriate macro-actions are identified with this method and discarded from agent choices. Experimental results show a significant improvement in results after pruning macro-actions.
A Flexible and Efficient Reconfigurable Architecture based on Multi-Agent Systems
2018 9th International Symposium on Telecommunications (IST), 2018
Nowadays, reconfigurable architectures (RAs) are very popular to design digital systems. One of t... more Nowadays, reconfigurable architectures (RAs) are very popular to design digital systems. One of the most important property of this system is flexibility. Runtime remapping is the concern that related to the flexibility concern. Recently some methods have been presented for runtime remapping. These methods aren’t reliable because a single agent or compiler remap the architecture. In this paper, we propose an intelligent reconfigurable architecture that its structure is based on a hierarchical multi-agent system. The method prepares runtime remapping for RA by interview and interaction between the system’s agents. We designed the structure in three layers. In this work, the aim of runtime remapping is enhancing flexibility and lifetime of the RA. For evaluation, we have simulated and implemented the method by HDL coding to study its feasibility and practicality. We evaluated the RA by three benchmarks. Results show the effect of our method on the lifetime and flexibility of the syste...
We investigate the prevalence rate of smoking in Covid-19 patients and examine whether there is a... more We investigate the prevalence rate of smoking in Covid-19 patients and examine whether there is a difference in the distribution of smokers between the two statistical populations of critically ill patients with Covid-19 and the entire Iranian population or not. To do this, we first prepared a sample of 1040 Covid-19 patients admitted to hospitals in Tehran, Rasht, and Bojnord. Then, through the non-parametric statistical runs test, we show that the sample was randomly selected, and it is possible to generalize the result of tests on the sample to the community of hospitalized Covid-19 patients. In continuation, we examined the hypothesis that the smoking prevalence among Covid-19 patients admitted to hospitals is equal to the prevalence rate of smoking in Iranian society. For this purpose, we used the non-parametric chi-square test, and it was observed that this hypothesis is rejected. The data show a significant difference in the prevalence of smoking between critically ill Covid-...
Extracting Bottlenecks Using Object Recognition in Reinforcement Learning
Developing and Evaluating A Real time and Energy Efficient Architecture for An Internet of Health Things
2020 4th International Conference on Smart City, Internet of Things and Applications (SCIOT), 2020
Real-time health monitoring systems play a critical role in preventing heart disease by processin... more Real-time health monitoring systems play a critical role in preventing heart disease by processing vital sign monitoring data. The internet of things will enhance the entire health care service-delivery and can lead to reducing the immediate risk in real-time. Fog assisted health-care IoT system and Edge computing technology are an emerging paradigm that reducing emergency response time and enhanced the quality of experiences. In hierarchical architectures, the task placement strategy for a task or service is a challenge that requires a successful response. In this paper, we propose a quality of experiences-aware health monitoring application according to patients’ demands and evaluate to clarify the impacts of different scenarios. The appropriate task placement scenarios in this hierarchical architecture lead to a satisfying quality of experience for improving health services. These policies are evaluated in terms of health services delivery, energy consumption, cost of execution(CoE) in the cloud, and network usage by iFogSim toolkit. In this paper, we have developed a new approach to improve resource management policies to achieve satisfactory results.
Journal of Intelligent Manufacturing, 2020
This paper proposes a novel incremental model for acquiring skills and using them in Intrinsicall... more This paper proposes a novel incremental model for acquiring skills and using them in Intrinsically Motivated Reinforcement Learning (IMRL). In this model, the learning process is divided into two phases. In the first phase, the agent explores the environment and acquires task-independent skills by using different intrinsic motivation mechanisms. We present two intrinsic motivation factors for acquiring skills by detecting states that can lead to other states (being a cause) and by detecting states that help the agent to transition to a different region (discounted relative novelty). In the second phase, the agent evaluates the acquired skills to find suitable ones for accomplishing a specific task. Despite the importance of assessing task-independent skills to perform a task, the idea of evaluating skills and pruning them has not been considered in IMRL literature. In this article, two methods are presented for evaluating previously learned skills based on the value function of the assigned task. Using such a two-phase learning model and the skill evaluation capability helps the agent to acquire taskindependent skills that can be transferred to other similar tasks. Experimental results in four domains show that the proposed method significantly increases learning speed.
Education and Information Technologies, 2018
Gaining the attention is the first key step to enhance learning. In Attention-Deficit/Hyperactivi... more Gaining the attention is the first key step to enhance learning. In Attention-Deficit/Hyperactivity Disorder (ADHD) as the most prevalent deficit in school age, the learners face some impairment in attention that requires appropriate intervention. An environment that embedded Pedagogical Agent in computer-assisted instruction (CAI) has been designed to support learning through gaining and guiding attention to relevant information for these students. This study investigated how much the presence of pedagogical agent can improve learning in ADHD students. The learning environment was integrated with a pedagogical agent, named Koosha, as a tutor and motivator. This study employed a pretest and posttest experimental design with a control group. The statistical population was 30 boy students with ADHD in primary school from the North of Iran. The participants were randomly assigned to work with either an agent
International Journal of Network Management, 2017
SummarySession initiation protocol (SIP) is a widely used standard protocol for multimedia applic... more SummarySession initiation protocol (SIP) is a widely used standard protocol for multimedia applications and IP multimedia subsystems. Internet protocol multimedia subsystem was introduced by the 3GGP signaling foundation as part of a set of next generation network architectures. Despite having useful practical features, SIP does not have suitable mechanisms to handle overload. This challenge creates a sharp drop in quality of service for next generation network users. Because a distributed SIP network is a complex system composed of subsystems interacting with one another, multiagent system is an appropriate method to model network interactions and to solve overload in SIP networks. In this paper, holonic organization is applied to reduce the multiagent system complexity in modeling a large SIP network. Holonic organization is a hierarchical structure in which each holon covers a geographical area of the SIP network at the first level. At the second level, upper‐level holons control...
Design and Implementation of a Fuzzy Control System for a ST2 Supersonic Wind Tunnel Nozzle
An automatically controlled variable converging-diverging supersonic nozzle with high dependabili... more An automatically controlled variable converging-diverging supersonic nozzle with high dependability is required for wide range of Mach numbers in a supersonic wind tunnel. In this work, the performance of a converging-diverging nozzle of the ST2 supersonic wind tunnel was evaluated and analyzed. Based on our results, an automatic control system for adjustment of the nozzle was designed and implemented. Then, a fuzzy control system was used to optimize the wind tunnel's performance and behavier. The results show that, it is supprior in speed, precision, stability, and reliability. In addition, the new system has caused better wind tunnel overall performance, enhancing its reliability and reducing its cost.
Real time prediction of time delays in a networked control system
2008 3rd International Symposium on Communications, Control and Signal Processing, 2008
Page 1. Behrooz Rahmani dept. of Mechanical Engineering IUST Tehran, Iran [email protected] Am... more Page 1. Behrooz Rahmani dept. of Mechanical Engineering IUST Tehran, Iran [email protected] Amir HD Markazi dept. of Mechanical Engineering IUST Tehran, Iran ir.ac.iust @ markazi Naser mozayani dept. of Computer ...
7'th International Symposium on Telecommunications (IST'2014), 2014
Communities are basic units of complex networks and understanding of their structure help us to u... more Communities are basic units of complex networks and understanding of their structure help us to understand the structure of a network. Communities are groups of nodes that have many links inside and few links outside them. Community detection in a network can be modeled as an optimization problem. We can use some measures such as Modularity and Community Score for evaluating the quality of a partition of nodes. In this paper, we present a new algorithm for detecting communities in networks based on an Estimation of Distribution Algorithm (EDA) with the assumption that the problem variables are independent. EDAs are those evolutionary algorithms that build and sample the probabilistic models of selected solutions instead of using crossover and mutation operators. In this paper, we assess our algorithm by synthetic and real data sets and compare it with other community detection algorithms.
The Study of Handover in Mobile IPv6 Networks
2008 International Conference on Advanced Computer Theory and Engineering, 2008
Abstract Mobile IP protocol (MIP) is a standard protocol that allows users to maintain non-stop c... more Abstract Mobile IP protocol (MIP) is a standard protocol that allows users to maintain non-stop connectivity with their home address regardless of physical movement. It would cover the significant of this solution on the industry's wireless infrastructure, its impact on users, ...
Procedia - Social and Behavioral Sciences, 2013
Dropout rate between 20 to 80 percent has been reported in e-learning, so decreasing dropout rate... more Dropout rate between 20 to 80 percent has been reported in e-learning, so decreasing dropout rate is one of the major challenges of e-learning systems. The aim of this study is to identify the theories that explain the success rate of elearners. We used a quantitative content analysis by reviewing the findings of 24 major studies in this field. Findings revealed that motivational theories(f:13); self-regulated learning(f:6) and interaction(f:5) are the most important explanatory theories for elearner success. Results from 223 elearner at IUST elearning center showed that there are relationship between self-regulation and elearner dropout, in addition the results of t-test revealed that persistence elearner (M=3.50,SD=.66) had significantly high self-regulatory score than the dropout group (M=3.24,SD=.80), t=-2.54(221),p=.01.
Uploads
Papers by Nasser Mozayani