فهرست مطالب کامبیز بدیع

Interpreting contents in social networks with the aim of analyzing the sentiment of their narrators is of particular significance. In this paper, we present a framework for such a purpose, which is able to classify the messages hidden in contents based on using some rule-type protocols with high abstraction level. According to this framework, items such as prosodic of a content's narrator, context of disseminating a content and the key propositions in a content's text are regarded in the condition part of a protocol, while the possible classes for the message in a content are considered as its action part. It is to be noted that the proposed rule-type protocols can equally be used for other languages due to the generic-ness of the above-mentioned items. Results of computer simulations on a variety of different contents in the social networks show that the proposed framework is sufficiently capable of analyzing the sentiment of the contents' narrators in these networks.

Lung cancer is a highly serious illness, and detecting cancer cells early significantly enhances patients' chances of recovery. Doctors regularly examine a large number of CT scan images, which can lead to fatigue and errors. Therefore, there is a need to create a tool that can automatically detect and classify lung nodules in their early stages. Computer-aided diagnosis systems, often employing image processing and machine learning techniques, assist radiologists in identifying and categorizing these nodules. Previous studies have often used complex models or pre-trained networks that demand significant computational power and a long time to execute. Our goal is to achieve accurate diagnosis without the need for extensive computational resources. We introduce a simple convolutional neural network with only two convolution layers, capable of accurately classifying nodules without requiring advanced computing capabilities. We conducted training and validation on two datasets, LIDC-IDRI and LUNA16, achieving impressive accuracies of 99.7% and 97.52%, respectively. These results demonstrate the superior accuracy of our proposed model compared to state-of-the-art research papers.

Nowadays, learning classifier systems have received attention in various applications in robotics, such as sensory robots, humanoid robots, intelligent rescue and rescue systems, and control of physical robots in discrete and continuous environments. Usually, the combination of an evolutionary algorithm or intuitive methods with a learning process is used to search the space of existing rules in assigning the appropriate action of a category. The important challenge to increase the speed and accuracy in reaching the goal in the maze problems is to use and choose the action that the stimulus is placed on the right path instead of repeatedly hitting the surrounding obstacles. For this purpose, in this article, an intelligent learning classifier algorithm of accuracy-based learning classifier systems (XCS) based on limited memory is used, which according to the input and actions applied to the environment and the reaction of the stimulus, the rules It is optimally identified and added as a new classifier set to the accuracy-based learning classifier systems (XCS) algorithm in the next steps. Among the achievements of this method, it can be based on reducing the number of necessary steps and increasing the speed of reaching the stimulus to the target compared to the accuracy-based learning classifier systems (XCS) algorithm.

Recently, learning classifier systems are used to control physical robots, sensory robots, and intelligent rescue systems. The most important challenge in these systems, which are models of real environments, is its non-markov quality. Therefore, it is necessary to use memory to store system states in order to make decisions based on a chain of previous states. In this research, a memory-based XCS is proposed to help use more effective rules in classifier by identifying efficient rules. The proposed model was implemented on five important maze maps and led to a reduction in the number of steps to reach the goal and also an increase in the number of successes in reaching the goal in these maps.

Within the past years, probability and possibility have both been elaborated as two key concepts in formulating the issue of uncertainty in various domains of decision making like pattern recognition. Possibility pays attention to the existing structure of a pattern (or a situation), while the main concern of probability is to emphasize on the problems experienced in the past. When we talk about "structure" the question is how the meaningful components in a pattern (or a situation) may come into existence based on specific models. Accordingly the perception based on possibility takes us to the paint where a set of possibilities with regard to these meaningful components come together to figure out the class identity of the related pattern based on the possibility of the presence of these specific models. In this paper, we try to show how the possibility of a pattern class can be assessed on the ground of the information belonging to its meaningful components which are supposed to come into existence based on specific models. To give a clear image of the proposed approach, recognition of shape patterns that are structural in nature, is taken into consideration.

Recently, network representation has attracted many research works mostly concentrating on representing of nodes in a dense low-dimensional vector. There exist some network embedding methods focusing only on the node structure and some others considering the content information within the nodes. In this paper, we propose HDNR; a hybrid deep network representation model, which uses a triplet deep neural network architecture that considers both the node structure and content information for network representation. In addition, the author's writing style is also considered as a significant feature in the node content information. Inspired by the application of deep learning in natural language processing, our model utilizes a deep random walk method to exploit inter-node structures and two deep sequence prediction methods to extract nodes' content information. The embedding vectors generated in this manner were shown to have the ability of boosting each other for learning optimal node representation, detecting more informative features and ultimately a better community detection. The experimental results confirm the effectiveness of this model for network representation compared to other baseline methods.