• Home
  • behrooz masoumi
  • OpenAccess
    • List of Articles behrooz masoumi

      • Open Access Article

        1 - Ensemble Learning Improvement through Reinforcement Learning Idea
        Mohammad Savargiv Behrooz Masoumi Mohammadreza Keyvanpor
        Ensemble learning is one of the learning methods to create a strong classifier through the integration of basic classifiers that includes the benefits of all of them. Meanwhile, weighting classifiers in the ensemble learning approach is a major challenge. This challenge More
        Ensemble learning is one of the learning methods to create a strong classifier through the integration of basic classifiers that includes the benefits of all of them. Meanwhile, weighting classifiers in the ensemble learning approach is a major challenge. This challenge arises from the fact that in ensemble learning all constructor classifiers are considered to be at the same level of distinguishing ability. While in different problem situations and especially in dynamic environments, the performance of base learners is affected by the problem space and data behavior. The solutions that have been presented in the subject literature assumed that problem space condition is permanent and static. While for each entry in real, the situation has changed and a completely dynamic environment is created. In this paper, a method based on the reinforcement learning idea is proposed to modify the weight of the base learners in the ensemble according to problem space dynamically. The proposed method is based on receiving feedback from the environment and therefore can adapt to the problem space. In the proposed method, learning automata is used to receive feedback from the environment and perform appropriate actions. Sentiment analysis has been selected as a case study to evaluate the proposed method. The diversity of data behavior in sentiment analysis is very high and it creates an environment with dynamic data behavior. The results of the evaluation on six different datasets and the ranking of different values of learning automata parameters reveal a significant difference between the efficiency of the proposed method and the ensemble learning literature. Manuscript profile
      • Open Access Article

        2 - An Improved Real-Time Noise Removal Method in Video StreamBased on Pipe-and-Filter Architecture
        Vahid Fazel Asl Babak Karasfi Behrooz Masoumi Mohamadreza Keyvanpor
        Automated analysis of video scenes requires the separation of moving objects from the background environment, which could not separate moving items from the background in the presence of noise. This paper presents a method to solve this challenge; this method uses the D More
        Automated analysis of video scenes requires the separation of moving objects from the background environment, which could not separate moving items from the background in the presence of noise. This paper presents a method to solve this challenge; this method uses the Directshow framework based on the pipe-and-filter architecture. This framework trace in three ways. In the first step, the values of the MSE, SNR, and PSNR criteria calculate. In this step, the results of the error criteria are compared with applying salt and pepper and Gaussian noise to images and then applying median, Gaussian, and Directshow filters. In the second step, the processing time for each method check in case of using median, Gaussian, and Directshow filter, and it will result that the used method in the article has high performance for real-time computing. In the third step, error criteria of foreground image check in the presence or absence of the Directshow filter. In the pipe-and-filter architecture, because filters can work asynchronously; as a result, it can boost the frame rate process, and the Directshow framework based on the pipe-and-filter architecture will remove the existing noise in the video at high speed. The results show that the used method is far superior to existing methods, and the calculated values for the MSE error criteria and the processing time decrease significantly. Using the Directshow, there are high values for the SNR and PSNR criteria, which indicate high-quality image restoration. By removing noise in the images, you could also separate moving objects from the background appropriately. Manuscript profile
      • Open Access Article

        3 - Utilizing Generalized Learning Automata for Finding Optimal Policies in MMDPs
        Samaneh Assar Behrooz Masoumi
        Multi agent Markov decision processes (MMDPs), as the generalization of Markov decision processes to the multi agent case, have long been used for modeling multi agent system and are used as a suitable framework for Multi agent Reinforcement Learning. In this paper, a g More
        Multi agent Markov decision processes (MMDPs), as the generalization of Markov decision processes to the multi agent case, have long been used for modeling multi agent system and are used as a suitable framework for Multi agent Reinforcement Learning. In this paper, a generalized learning automata based algorithm for finding optimal policies in MMDP is proposed. In the proposed algorithm, MMDP problem is described as a directed graph in which the nodes are the states of the problem, and the directed edges represent the actions that result in transition from one state to another. Each state of the environment is equipped with a generalized learning automaton whose actions are moving to different adjacent states of that state. Each agent moves from one state to another and tries to reach the goal state. In each state, the agent chooses its next transition with help of the generalized learning automaton in that state. The experimental results have shown that the proposed algorithm have better learning performance in terms of the speed of reaching the optimal policy as compared to existing learning algorithms. Manuscript profile
      • Open Access Article

        4 - A New Multi-Agent Bat Approach for Detecting Community Structure in Social Networks
        Saeed Alidoost Behrooz Masoumi
        The complex networks are widely used to demonstrate effective systems in the fields of biology and sociology. One of the most significant kinds of complex networks is social networks. With the growing use of such networks in our daily habits, the discovery of the hidden More
        The complex networks are widely used to demonstrate effective systems in the fields of biology and sociology. One of the most significant kinds of complex networks is social networks. With the growing use of such networks in our daily habits, the discovery of the hidden social structures in these networks is extremely valuable because of the perception and exploitation of their secret knowledge. The community structure is a great topological property of social networks, and the process to detect this structure is a challenging problem. In this paper, a new approach is proposed to detect non-overlapping community structure. The approach is based on multi-agents and the bat algorithm. The objective is to optimize the amount of modularity, which is one of the primary criteria for determining the quality of the detected communities. The results of the experiments show the proposed approach performs better than existing methods in terms of modularity. Manuscript profile
      • Open Access Article

        5 - Improving Agent Performance for Multi-Resource Negotiation Using Learning Automata and Case-Based Reasoning
        Monireh Haghighatjoo Behrooz Masoumi Mohamad Reza Meybodi
        In electronic commerce markets, agents often should acquire multiple resources to fulfil a high-level task. In order to attain such resources they need to compete with each other. In multi-agent environments, in which competition is involved, negotiation would be an int More
        In electronic commerce markets, agents often should acquire multiple resources to fulfil a high-level task. In order to attain such resources they need to compete with each other. In multi-agent environments, in which competition is involved, negotiation would be an interaction between agents in order to reach an agreement on resource allocation and to be coordinated with each other.In recent years, negotiation has been employed to allocate resources in multi-agent systems. Yet, in most of the conventional methods, negotiation is done without considering past experiments. In this paper, in order to use experiments of agents, a hybrid method is used which employed case-based reasoning andlearning automata in negotiation. In the proposed method, the buyer agent would determine its seller and its offered price based on the passed experiments and then an offer would be made. Afterwards, the seller would choose one of the allowed actions using learning automata. Results of the experiments indicated that the proposed algorithm has caused an improvement in some performance measures such as success rate. Manuscript profile