Sharif Digital Repository

As the core of this thesis, we express the relation between optimal transportation and convex geometry especially the variational approach to solve Alexandrov problem, which leads to a geometric interpretation to adversarial models and propose a novel framework for these models. Along the way, we peruse generative adversarial networks from optimal transportation view and show that generator calculates the transportation map and the discriminator computes Wasserestein distance, which is equivalent to Kantorovich potential. By using optimal mass transportation theory and choosing an especial cost function c, we see that the generator and discriminator are equivalent. Therefore, once the... 

Over the past two decades, the interest in exploring consumers' behavior has risen to popularity. Nowadays, it is necessary for marketers to reach out to the right consumers at the right time and with the right message. Even though the positive impact of tracking consumers throughout their purchase journey and its managerial implications have been emphasized several times by academics, There is still a lack of practical research on how to discern the consumers' stage in the purchase journey from a holistic point of view. In this study, we develop several machine learning models, including RNNs (GRU and LSTM), Transformers, and XGBoost, by utilizing historical data of Yektanet (the leading... 

Recently, the advent of online platforms as an intermediary tool has provided an opportunity for service providers and manufacturers to offer their services/products to consumers. Ride-hailing platforms are an example of these businesses where independent drivers and passengers are matched via an online platform. One of the most significant challenges faced by these companies is the emergence of strategic behaviors on both the demand and the supply side, which might affect the performance and profitability of the platform.In this thesis, we study the strategic behavior of drivers and its impact on their performance, using the data of one of the largest ride-hailing companies in Iran. We use... 

There are always some unpredictable phenomena happen in the process of constructing of a downstream petrochemichal plant resulting to delays that increase the cost compared with the benefits or sometimes cancelation of the projects that abodones the investors from the returns at all. On of the reasons that these issues can dwibdles the benefits is that most of Iranian’s downstream companies do not prepare suitable and sufficient capabilitities to cope with such a risks. In this thesis, at first, influensive risks on the downstream oil and gas projects in Iran have been studied based on a comprehensive litreture review, secondly using a questionnaire, these extracted risks have been assessed... 

This study explores topics discussed in positive eWOM of robotic and conventional hotels through a comprehensive analysis of online reviews using Latent Dirichlet Allocation (LDA). LDA is an advanced topic modeling technique. To uncover topics discussed in eWOM of robotic hotels and conventional hotels, and then compare them to see if there are any meaningful statistical differences between them, three separate LDA models were implemented on reviews of robotic hotels, conventional hotels, and the overall rereviews of all hotels, respectively. The results reveal that “service quality and overall atmosphere” is the most common topic discussed considering both hotel types, with key themes... 

This thesis examines the impact of electronic word-of-mouth (e-WOM) on product demand forecasting, specifically in the context of video game consoles. This study examines the importance of emotion-based topics and their impact on demand forecasting at two stages of the product life cycle: pre-launch and post-launch. Using sentiment analysis and topic modeling, this study uncovers product sales drivers and captures evolving consumer sentiment throughout the product lifecycle. By integrating insights from diffusion theory and consumer information search theory, this research contributes to the field of demand forecasting using social media data. This research shows that by using the topics... 

This research examines the attributes, themes, and topics discussed in electronic word-of-mouth (e-WOM) generated in TripAdvisor for robotic hotels in comparison to traditional ones. The first objective is to explore the themes and topics within TripAdvisor negative e-WOMs for both types of hotels and analyze how these topics differ between the two types of hotels, using the Latent Dirichlet Allocation (LDA) method. The research shows that which topics are mostly discussed in TripAdvisor negative e-WOMs for both types of hotels. Additionally, the research explores whether different themes, topics, and attributes have varying effects on user ratings in traditional hotels versus robotic ones,... 

Residential satisfaction plays a pivotal role in shaping liveable cities, profoundly influencing individuals' health and well-being. In contrast, dissatisfaction in living environments often triggers relocation intentions, leading to significant vacancy costs for landlords and property managers. While previous studies predominantly relied on traditional methods like surveys and interviews, our research adopts a novel, bottom-up approach, analysing this phenomenon through the lens of residents themselves. Leveraging a custom-developed Python-based web scraper, we meticulously gathered 195,000 online reviews from apartments.com. This comprehensive dataset spans the most populous cities and... 

In this thesis, a field effect transistor (FET) using armchair graphene nanoribbon as the channel is simulated, and the effects of changing nanoribbon width and length, as well as adding defects, are also studied. To obtain the Hamiltonian matrix and the energy band structure of graphene nanoribbon, tight binding method is used in which the first and third neighbor approximation is considered. Also, to maximize accuracy, we also considered the edge bond reconstruction. To obtain the transport characteristics of the transistor, such as the transmission coefficient and the density of states (DOS), Poisson and Schrodinger equations are solved self-consistently. We used the nonequilibrium... 

Mobile ad hoc networks (MANETs) are networks which consist entirely of mobile nodes, placed together in ad hoc manner, i.e. with minimal prior planning. In these networks, all nodes have routing capabilities and forward data packets for other nodes. Nodes can enter or leave the network at any time and may also be mobile. Hence, the network topology changes frequently. There are lots of challenges in these networks, which make routing to be a hard task. These challenges arise from the dynamic and unplanned nature of these networks such as unreliability of wireless communication, limited resources available in terms of bandwidth, processing capacity, network size, and etc. Due to these... 

One of the most intrinsic challenges of flash-based Solid State Drives (SSDs) is erasebefore-write limitation and the limited endurance of flash chips. Wear leveling and garbage collection are two mechanisms implemented in SSD’s controller to enhance endurance and performance. While wear leveling attempts to distribute erasures across all blocks in an even manner, it imposes a new security challenge on SSD,which leads to the presence of invalid data blocks in flash chips. Therefore, some data blocks that are logically deleted by the user are still available in flash chips and can be recovered by software or hardware recovery tools.
In this paper, a new criteria named vulnerability time... 

This thesis presents the results of an experimental and analytical study on the application of the strapping technique for seismic retrofit of reinforced concrete columns. In this technique, the columns are externally confined by post-tensioned metal strips. In the first phase of the experiments, the compressive stress-strain behavior of strapped cylindrical and prismatic specimens was studied. Effects of several parameters were studied including shape and size of specimens, strength of plain concrete, mechanical characteristics of confining strips and the level of pre-tensioning of the strips. The second phase of the tests included studying the lateral performance of strapped RC columns.... 

The spine like every other mechanical pillar, is exposed to buckling and loss of stability. While existing biomechanical models emphasize the pressure force on the disk as the main cause of injury, there is also a possibility of local buckling phenomenon in vertebral discs. Because of the prevalence and high cost of lower back pain, it is essential to evaluate the forces carried by disks and lumbar muscles during occupational activities more accurately. In this regard, hybrid EMG-assisted optimization (EMGAO) approaches are most common methods for estimation of spinal loads. These models, not only use EMG data to be physiologically creditable, but also satisfy equilibrium requirements at all... 

Polychlorinated biphenyls are a group of hydrocarbons that are composed of two benzene rings and chlorine admixture on it. Due to their special properties, these materials were added to oils for years to improve their properties. These compounds have many industrial applications in various industries, such as the production of fluorescent lamps, adhesives, dyes, and especially as a cooling and insulating liquid as the Askarel oil in power transformers. Due to the properties of these compounds, which include stability and bioaccumulation in fatty tissues, the wide use of these substances causes environmental contamination and leads to a variety of cancerous diseases. Among the cleaning...