With financial support from the Center for International Scientific Studies and Collaboration, the research project titled “Decentralized Algorithms (Parallel Processing) for Machine Learning, Artificial Intelligence, and Cyber-Physical Networks” has been carried out under a contract between Semnan University and the Center, supervised by Dr. Mohammadreza Doost Mohammadian.
In the rapidly growing research areas of Artificial Intelligence (AI) and Cyber-Physical Systems (CPS), the demand for fast, efficient, scalable, and resilient solutions has significantly increased in recent years. Traditional centralized methods in machine learning, control, optimization, and filtering face challenges such as managing big data and the complexity of existing systems, especially in applications like the Internet of Things (IoT). With the advancement of technologies and expansion into domains such as smart cities, autonomous vehicles, smart energy grids, and green computing, the importance of distributed and decentralized algorithms in shaping the future of AI and CPS is evident from multiple perspectives. These algorithms promise a fundamental transformation in how intelligent systems are understood, developed, and deployed.
First, the massive volume of data generated by interconnected devices and sensors necessitates distributed frameworks for processing and learning. Centralized architectures are unable to efficiently handle extensive information exchange, leading to processing delays and reduced quality of real-time decision-making. Distributed algorithms offer a decentralized solution with parallel processing capabilities, enabling the integration of machine learning models and data filtering across a network of nodes or agents connected to cloud platforms, thereby facilitating parallel processing and accelerating decision-making.
A video clip showcasing this achievement is available at the following link:
