Sharif Digital Repository

Demands on wireless access networks are growing enormously, thanks to deploying new generations of wireless communication systems—such as LTE, 4G, and in the near future 5G—and especially promoting new technologies, like Internet-of-Things (IoT). With the help of visible light communication (VLC) systems, networks have been developed the capacities of which can go beyond some Mbps, but this is not satisfying enough. Therefore, a newer version of VLC systems has been investigated, called Light-Fidelity (LiFi), which is complement to WiFi (in radio frequency networks(. Two main purposes of LiFi networks are obtaining a high-capacity wireless access network and using them for illumination... 

The utilization of tower cranes at construction sites presents numerous inherent risks. These cranes are commonly employed for lifting heavy loads, which carry the potential hazard of accidental falls. Simultaneously, workers may inadvertently overlook overhead dangers while focusing on their tasks. To mitigate these risks, laws in many countries explicitly prohibit individuals from occupying the vicinity directly beneath suspended loads, known as the fall zone. Such measures are vital to safeguard against the peril of heavy loads plummeting onto people. However, existing studies have not offered a comprehensive and efficient approach to identify crane load fall zone. To address this gap,... 

Sparse representation has been an important problem in recent decade. The main idea in this problem is that natural signals have information contents much lower than their ambient dimensions and as such, they can be represented by using only a few atoms. For example, if the dimension of signal is n, the purpose in sparse representation is to achieve the representation of signal in terms of s atom (s ≪ n). In sparse coding, the dictionary depends on the used signal. In some of the problem, dictionary is specified and sparse representation is obtained by this dictionary. In this case, because the dictionary is known, maybe sparse representation is not suitable for this signal. For this reason,... 

Due to the technological improvement, decreasing transistor’s size and increasing demand for more processing abilities, designing and implementing multi core processors becomes a hot topic. One of the direct effects of decreasing the size is increasing the rate of power consumption in a unit area. Thus, the generated heat by the processor will be increased. Increasing the temperature could has undesired effects on the other features of the processor such as reliability, efciency, and failure rate. The problems that caused due to the high temperature in the multicore processors is not limited to the maximum temperature, but temperature fluctuations and variation between different parts of the... 

Distance Metric Learning algorithms have been widely used in Machine Learning methods recently. In these algorithms a distance function between objecs (data points) is learned based on their labels or similarity and dissimilarity constraints. Recent works have shown that a good precision is obtained in classification or clustering methods which use these functions. Since in the current systems many of data points do not exist at the beginning and are added to the training set as the algorithm is run, online methods are needed to update learned metric due to new data.
In this thesis, we proposed a new online distance metric learning method that has higher performance than existing... 

The main goal of this research is to investigate different types of learning methods used on repetitive games. A repetitive game is a model for describing all kinds of frequent activities among humans or intelligent machines. Applying learning models to repeated games is the intersection of game theory and machine learning fields. In the field of game theory, this research is explored under the title of behavioral game theory, where the purpose is to predict the behavior of human beings in repeated games in static environments (without states) and study how they select their actions. In the field of machine learning, repeated games are referred to as multi-agent systems and include problems... 

Colorectal cancer (CRC) is one of the most common cause of cancer death worldwide. Identification of pivot genes in colorectal cancer can play an important role as biomarkers in predicting and early diagnosis and reducing the number of deaths caused by this disease. In this study, the aim of which is to discover pivot genes in colorectal cancer, six microarray datasets selected from the GEO database including 277 tumor tissue samples and 325 normal colon tissue samples. After data processing, differentially expressed genes and CRC-related genes were screened and 285 shared genes between them were identified for subsequent analysis. Based on 285 shared genes, the protein-protein interaction... 

Buildings account for more than 30% of the world’s total energy consumption. Among building end-uses, air conditioning and in particular cooling systems have a major share of more than 50%. Therefore, design of optimal controllers for AC systems has become increasingly important. Classical and model-free control methods typically lack the ability to optimize energy consumption. On the other hand, model-based optimal control methods rely on precise modeling, which is difficult to acquire due to the complexity of the AC system dynamics.In recent years, deep reinforcement learning has become a popular choice for optimal control of systems with complex dynamics. In this thesis, a deep... 

One way for businesses to grow and compete, in any age (especially the digital age), is to create a Brand Relevance through creating or finding, and then owning new categories or subcategories. In this way, instead of beating competitors; they become irrelevant by enticing customers to buy a new category or subcategory for which other alternative brands are not considered relevant. Firms traditionally rely on interviews and focus groups to identify these subcategories and customer needs. Nowadays, with the growth of social media, user-generated content (UGC) is also a good alternative source. However, Due to the large size of UGC and the non-informative or repetitive data it contains,... 

Reservoir Computing is a novel computing paradigm that uses a nonlinear recurrent dynamical system to carry out information processing. Recent electronic and optoelectronic Reservoir Computers based on an architecture with a single nonlinear node and a delay loop have shown performance on standardized tasks comparable to state-of-the-art digital implementations. Here we report an all-optical implementation of a Reservoir Computer, made of off-the-shelf components for optical telecommunications. It uses a semiconductor optical amplifier as nonlinearity, and a Fabry-Perot Resonator as a key element to establish the virtual nodes, connecting them and consequently, build the virtual neural... 

In recent years, adversarial attacks have become a severe challenge in the field of security and stability of neural networks. These attacks in the field of medical images are able to cause misdiagnosis of the network, while hostile samples are considered normal from the perspective of a human observer. There are different ways to deal with these attacks. These methods mainly require full access to the neural network and knowledge of the type of hostile attack, as a result, the performance of these defense methods against unknown attacks drops drastically. In this research, we aim to find a method for identifying adversarial samples without knowing the architecture of the target neural... 

Increasing in insider threats of the organization, made it necessary to use a security solution to investigate anomalies within the organization. Diagnosis of insider anomalies is based on examining the behavior of any entity, whether employees or systems of the organization. Entities are classified into different categories based on reported events. The high consumption of resources, the creation or elimination of entities, and the variability of their behavior over time are among the major challenges in diagnosing insider anomalies. In this research, while defining the process of diagnosing insider anomalies, a possible solution is presented considering the above challenges. In the... 

The influence mechanisms of executive/non-executive board compostion on board’s attribution style have been explored. In two case studies, I have examined boards’ causal interpretation about a critical incident which had a significant impact on firm performance. A qualitative research approach has been taken in the case studies. Data was gathered through semi-structured interviews with all board members and was analyzied with the theme analysis method. According to the results, board of directors’ collective attribution formation process can be devided into three stages of members’ expression of information and facts at board meetings, forming the individual judgements of members and at... 

After three decades of research, the concept of emotion as an aspect of initelligence in artificial systems, is still unclear. Most of emotion models in AI are aimed to simulate the emotional states of human in differenct levels of abstraction and functionality. Therefore, the concept of emotions in artificial systems is completely dependent on human being’s emotional states and can not be used as a general tool for improving the intellingence of artificial systems. In this thesis, we have tried to show that defining emotion in a more abstract level can expand the domain of its appications in AI. According to some existing theories about emotions, we modeled emotions as local semi-soloutions... 

We present a deep neural network to classify dermatological disease from patient images. Using self-supervised learning method we have utilized large amount of unlabeled data. We have pre-trained our model on 27000 dermoscopic images gathered from razi hospital, the best dermatological hospital in Iran, along with 33000 images from ISIC 2020 dataset. We have evaluated our model performance in semi-supervised and transfer learning approaches. Our experiments show that using this approach can improve model accuracy and PRC up to 20 percent on semi-supervised setting. The results also show that pretraining can improve classification PRC up to 20 percent on transfer learning task on HAM10000... 

Sample efficiency and systematic generalization are two long-standing challenges in sequential decision-making problems, especially, in reinforcement learning settings. It is hypothesized that involving natural language in conjunction with other observation modalities in decision-making environments can improve generalization due to its compositional and open-ended nature, and sample efficiency due to the concise information summarized in relatively short linguistic units. By exploiting this information and the compositional structure of the language, one can achieve an abstract and factored understanding of the environment and the task at hand. To do so, it is necessary to find the proper... 

Over the past few decades, with the exponential increase in data volume, scientists and researchers have tried to discover relationships and algorithms for productivity and find useful information from this amount of data in various fields. Their efforts in data analysis have led to the development of algorithms in the big data field. The result of researchers' working in multiple fields has come to aid the people of science and technology. Among the most important of these areas, we can mention the health and medical sectors, financial sectors, services, manufacturing sectors, etc. The purpose of this study is to enter the financial industry and use data mining tools. One of the newest and... 

Expert finding in bibliographic networks has received increasing attention in recent years. This task concerns with finding relevant researchers for a given topic. In this thesis, we propose a model to determine authority of authors who have participated in the Communities. This model has a little improvement over community based baseline model. However, due to the low performance of community based models, the proposed authority based model cannot improve the document based baseline models either. Therefore, we try to improve document based models, instead of community based models and have proposed two other models which are based on authors’ topic dominance for expert finding. Document... 

This thesis aims at reducing the need for labeled data in the SLU domain by the means of active Learning methods. This need is due to the lack of labeled datasets for Spoken Language Understanding (SLU) in the Persian language, and fairly high labeling costs. Active learning methods enables the learner to choose the most informative instances to be labeled and used for training, and prevents labeling uninformative or redundant instances. For modeling the SLU system, several statistical models namely MLN (Markov Logic Networks), CRF (Conditional Random Fields), HMM (Hidden Markov Model) and HVS (Hidden Vector State) were reviewed, and finally CRF was chosen for its superior performance. The... 

SDN-based cloud is created by new thechnologies. This infrastructure is more programmable, manageable and configurable. However SDN-based cloud is vulnerable to the DDoS attacks. A lot of researches has been accomplished to prevent these kind of attacks. Solutions that proposed in these papers are based on machine learning, statistical analysis of traffic or combination of these approaches. In this research an efficient method has been introduced, for detecting DDOS attack in SDN-based cloud environment. Detection system is based on extreme learning machine (ELM). ELMs has been pruned with genetic algorithm (GAP-ELM). Detection of attack in the proposed system, has been accomplished with...