جستجوی مقالات مرتبط با کلیدواژه « یادگیری » در نشریات گروه « برق »

Towards realizing an intelligent networked society, enabling low-cost low-energy connectivity for things, also known as the Internet of Things (IoT), is of crucial importance. While the existing wireless access networks require centralized signaling for managing network resources, this approach is of less interest for future generations of wireless networks due to the energy consumption in such signaling and the expected increase in the number of IoT devices. Then, in this work, we investigate leveraging machine learning for distributed control of IoT communications. Towards this end, first, we investigate low-complex learning schemes that are applicable to resource-constrained IoT communications. Then, we propose a lightweight learning scheme which enables the IoT devices to adapt their communication parameters to the environment. Further, we investigate analytical expressions presenting the performance of a centralized control scheme for adapting communication parameters of IoT devices and compare the results with the results from the proposed distributed learning approach. The simulation results confirm that the reliability and energy efficiency of IoT communications could be significantly improved by leveraging the proposed learning approach.

In addition to initial roles in the form of operational readiness for defensive battles, military organizations should be able to explore new ways and technologies for one to one with regard to the dynamic environment they face. This requires careful strategic planning and simultaneous use of the two capabilities of exploration and exploitation in these organizations. Strategic dynamics in such a situation require organizational learning in both dimensions of exploration and for profit. According to the importance of strategy field in defensive organizations, this study focuses on a conceptual explanation for ambivalent strategic learning in the air defense Headquarters Khatam-All-Anbiya and is trying to explain the concept of ambivalent strategic learning. The method of doing qualitative research was used to collect data from library and field method (interviews with experts in research field). The research community of experts in air defense Headquarters Khatam-All-Anbiya and faculty members of universities in Tehran are related to the background of research in the field of learning and education.In order to select the experts for interviews and identify components, the purposeful sampling method was used.The Questionnaire was completed with the help of guidance professors and consultants and interviews conducted to complete interviews and saturation of the interviews with the presence of experts and then with the content analysis method and the three coding stages, the strategic learning pattern was developed which includes components and strategic learning processes of ambivalent. It can serve as a basis for the design of education and learning in military centers.

In this paper, the dynamic Bayesian network for modeling behavior has been used for door-crossing behavior. This network easily trains and it supports dependencies between of data’s sensors and the desired behavior of robot on ideal environment and it detects error on data’ sensor by using data based method in real environment. Elimination of undesired sensor performs with structure changing base on interpretation of sensor’s data. Evaluation is based on robots with artificial error in the readings of the sensors. Proposed model for the "door-crossing", which is based on dynamic Bayesian networks, has been able to resist much higher than the model based on Bayesian networks (fixed structure), in the face of adverse and multiple interpretations of the constant error, drift, also cross-talk, random movement and slight changes in the environment map. This model is much resistant by adding maneuver part to it on dynamic environments. The performance of the proposed model on the simulation environment has been evaluated.

In recent years, many efforts have been done to design ontology learning methods and automate ontology construction process. The ontology construction process is a time-consuming and costly procedure for almost all domains/applications, so automating this process is a solution to overcome the knowledge acquisition bottleneck in information systems and reduce the construction cost. In this article a novel intelligent ontology learning method is proposed which can be used in many domains and applications. The proposed learning system has no need for initial common or specialized input ontologies or predefined semantic terms; indeed, the initial database anonly consists of input texts sets. The proposed learning system could extract associated ontologies of various domains using combined methods. To do this, a combination of linguistic, statisticaland machine learning methods based on the C-value method, the TF-IDF one, the neural network, and co occurance analysis are applied. So, first domain-related documents were collected. Then natural language processing methods such as C-value method were implemented for extracting meaningful terms from documents. Next, ART (Adaptive Resonance Theory) neural network was used to cluster documents and associated weight of terms was calculated by TF–IDF method in order to find candidate keyword for each cluster. Finally, co-occurrence analysis was used to construct concept hierarchy and complete the ontology. Results show that the proposed ontology learning method has a high precision comparing to similar studies.

Artificial Intelligence (AI) Techniques (such as learning) are used widely in agent-based systems. However, current research does not address a software engineering view on these techniques that support all the software development process. In this paper, we focus on requirement analysis – as the first step of the software development process and present techniques and tools to cover this shortage. In this regard, we provide a set of stable analysis patterns for learning capability of the agents. Stable analysis patterns are a set of meta-classes and their relations to analyze a specific issue in a domain-independent manner. Using stable analysis concepts, namely Enduring Business Themes (EBT), Business Objects (BO) and Industrial Objects (IO), these patterns represent the conceptual model of the learning. In this paper, we also apply these patterns on two case studies to investigate their applicability. These patterns are used as guidelines during analysis of learning. The main advantage of applying the stable analysis patterns in comparison with conventional analysis methods is modeling the knowledge of the learning analysis in addition to the ordinary classes of the domain. In addition, they generate more stable models via considering different levels of abstraction in the analysis.

In this paper a new structure based on Bayesian networks is presented to improve mobile robot behavior, in which there exist faulty robot sensors. If a robot likes to follow certain behavior in the environment to reach its goal, it must be capable of making inference and mapping based on prior knowledge and also should be capable of understanding its reactions on the environment over time. Old learning models for knowledge learning, especially on dynamic environment, are quite complex and have uncertainty in sensors. In this paper a new structure based on Bayesian network is presented for knowledge learning on robot behavior when the malfunction sensors exist. In this paper successful door crossing behavior is explained. In this issue the belt of ultrasonic sensors is used to receive environment information. Simulation results show that using the Bayesian network is very effective in robot behavior with faulty sensors.