arXiv subjects

Amin Mahmoudi

Publications and source records attributed to Amin Mahmoudi.

4 recordsLinked to original sources

arXiv2026

Temperature-Aware Scheduling of LLM Inference in Large-Scale Geo-Distributed Edge Data Centers with Distributed Optimization

The environmental impact of Large Language Models (LLMs) on data centers hosting these models is becoming a significant concern. While many efforts have focused on reducing the substantial training overhead of LLMs, carbon and water consumption during the inference phase can often surpass the costs associated with their training. The cooling systems of data centers are crucial in this context, but they are frequently modeled with a location-independent efficiency term. However, their energy efficiency is highly influenced by ambient temperature, which can vary significantly across different geographical locations. Leveraging this temperature diversity can help reduce total cooling energy costs and improve the performance of edge data centers. To address these critical sustainability issues related to LLMs, this study proposes a temperature-aware approach that co-optimizes LLM energy costs, carbon emissions, time-to-first token, and water consumption. The approach employs a distributed optimization algorithm based on an alternating direction method of multipliers, aimed at enhancing the sustainability of LLM hosting across geo-distributed edge data centers in Australia. Our method demonstrates reductions in cooling energy consumption and improves overall cost efficiency for geo-distributed cloud environments.

eess.SY↗

arXiv2026

Leveraging Quantum Annealing for Large-Scale Household Energy Scheduling with Hydrogen Storage

Hydrogen integration into microgrids facilitates the absorption of intermittencies from renewable energy resources. However, significant challenges remain due to complex optimization problems, particularly in large-scale applications involving multiple fuel cells (FCs) and electrolyzers (ELs) with numerous binary decision variables. This paper presents a hierarchical quantum annealing (QA) model predictive control-based power allocation framework aimed at accelerating these optimization problems. First, in a day-ahead stage, the framework determines the startup and shutdown of the FCs and ELs. The short-term stage then refines the output power of the FCs and the hydrogen generation rate of the ELs. The feasibility is evaluated through a case study consisting of multiple households in Australia. Our findings demonstrate that while the traditional optimization approach performs satisfactorily in scenarios with a small number of households, the QA approach becomes more appropriate and effectively solves the problem within an acceptable range as the number of connected households increases.

eess.SY↗

arXiv2021

Identifying Biased Users in Online Social Networks to Enhance the Accuracy of Sentiment Analysis: A User Behavior-Based Approach

The development of an automatic way to extract user opinions about products, movies, and foods from online social network (OSN) interactions is among the main interests of sentiment analysis and opinion mining studies. Existing approaches in the sentiment analysis domain mostly do not discriminate the sentences of different types of users, even though some users are always negative and some are always positive. Thus, finding a way to identify these two types of user is significant because their attitudes can change the analysis of user reviews of businesses and products. Due to the complexity of natural language processing, pure text mining methods may lead to misunderstandings about the exact nature of the sentiments expressed in review text. In this study, we propose a neural network classifier to predict the presence of biased users on the basis of users' psychological behaviors. The identification of the psychological behaviors of users allows us to find overly positive and overly negative users and to categorize these users' attitudes regardless of the content of their review texts. The experiment result indicates that the biased users can be predicted based on user behavior at an accuracy rate of 89%, 67% and 81% for three different datasets.

cs.SI↗

arXiv2021

User Behavior Discovery in the COVID-19 Era through the Sentiment Analysis of User Tweet Texts

The coronavirus disease (COVID-19) outbreak was declared a pandemic in March 2020 and since then it has had a significant effect on all aspects of life. Although we live in an information era, we do not have accurate information about this disease. Online social networks (OSNs) play a vital role in society, especially people who do not have trust in the government would tend to have more confidence in the evidence that is formed by social networks. The advantages of OSNs in the COVID-19 era are clear. For instance, social media enables people to connect with each other without the need for real-world face-to-face social interaction. Social media networks also act as a collective intelligence in the absence of world leadership. Therefore, in this study, considering the phenomenon of information diffusion in OSNs, we focus on the effects of COVID-19 on user sentiment and show the user behavior trend during the early months of the pandemic through mining and analyzing OSN data. Moreover, we propose a data-driven model to demonstrate how user sentiment changes over a period of time and how OSNs help us to obtain information on user behavior that is very important for the accurate prediction of future behavior. For this purpose, this study uses tweet texts about COVID-19 and the related network structure to extract significant features, and then presents a model attempting to provide a more comprehensive real picture of current and future conditions.

cs.SI↗