All news are also available on the Instagram channel @nesya_lab

Hyperdimensional Computing and Vector Symbolic Architectures for Automation and Design in Technology and Systems

Workshop of the Design, Automation and Test in Europe Conference (DATE 23)

This workshop will explore the interplay between different technologies and architectures that enable the automation of complex systems and enable design of efficient and effective solutions for real-world problems. The objective of this workshop is to bring together researchers in the fields of hyperdimensional computing, vector symbolic architectures, and automation and design in technology and systems to discuss the latest developments and challenges in these areas, and to identify synergies between them with an open discussion.

Call for Joining NESYA

Positions Available at the Laboratory

Join the NESYA laboratory! We're doing exciting research on electrical engineering, neural networks and quantum computing. We were born before the Italian PNRR and will also exist after the PNRR itself, because we aim at obtaining original and significant contributions with scientific results that last over time, going beyond any incidental or misleading funding.

There are open positions now for Master theses in Electronics, Telecommunications, Data Science, Artificial Intelligence and Management/Industrial Engineering. Next Fall we will be hiring for Ph.D. and Research Fellowship grants, as well as for professional collaborations in our forthcoming and brand new Startup.

No future unemployment, no faking positions. We rely on our strength only and on the resources that we can obtain thanks to our projects and our successes. It will be like this with you too!

Computational Intelligence in Electrical Systems

Special Issue on Energies (ISSN 1996-1073), an Open Access Journal by MDPI

Electrical systems play a central role in the energy transition from fossil fuels to renewables. This Special Issue is intended to bring forth advances in the use of computational intelligence tools (shallow and deep neural networks, fuzzy systems, evolutionary computation, etc.), in connection with statistical machine learning and signal processing techniques, for the solution of real-world problems related to electrical systems. Special attention should be paid to the distributed contexts of smart grids, renewable energy sources, energy storage systems, electric vehicle infrastructures, as well as to the energy/power aspects in ICT technologies and the related applications as, for instance, hungry data centers, green computing and green networking, EMC/EMI, energy harvesting, low-power micro/nano/optoelectronic systems, and so forth. Strategic tasks to be considered are pattern analysis, data regression and classification, optimization and control, decision-making, time series forecasting.

AI-based Time Series Analysis for Energy-related Applications and Systems

Special Session of the International Conference on Applied Intelligence and Informatics (AII 2022)

This session considers artificial intelligence and informatics applied to the energy-related fields and to those applications involving the “twin transition” of both digital and green/sustainable systems. A focus on time series analysis is expected, where the use of machine learning and deep learning techniques can lead to innovative solutions and/or to more efficient applications. Both people engaging theoretical foundations based on computational intelligence (i.e., neural networks, fuzzy logic, evolutionary computation, deep learning, etc.) or people using such computational techniques in practical applications (i.e., pattern recognition, data regression and classification, time series prediction, inverse modeling, multi-spectral image and data processing, sensor networks, smart grids, power delivery and control, characterization of structures and materials, and so forth) are advised to take part in this session.

Hyperdimensional Computing and Vector Symbolic Architectures in Neural Networks

Special Session of the IEEE World Congress on Computational Intelligence (IJCNN 2022)

This special session, which is hosted in the International Joint Conference on Neural Networks (IJCNN 2022) of WCCI 2022, is intended to bring forth advances in neural network architectures and applications via Hyperdimensional Computing (HDC) or Vector Symbolic Architectures (VSA). Submitted papers should propose novel algorithms or systems based on neural networks employing HDC/VSA with the goal of showing new solutions to supervised learning problems and applications or providing original insights on neural methods.

Variational Learning for Quantum Neural Networks

Special Session of the IEEE World Congress on Computational Intelligence (IJCNN 2022)

The main goal of this session, which is hosted in the International Joint Conference on Neural Networks (IJCNN 2022) of WCCI 2022, is to study the performances of quantum variational algorithms with respect to their classical counterpart, paving the way for a new paradigm of Computational Intelligence based on a hybrid quantum-classical approach. It is also important to investigate what could be the results of recurrent, convolutional, attention-based and all other neural models in a quantum environment, as well as their behavior and scalability when applied to real-world data through NISQ devices, by also addressing barren plateaus problems and exploring the possibilities of theorizing new ansatz models.

Randomization-based Fuzzy Neural Networks

Special Session of the IEEE World Congress on Computational Intelligence (FUZZ-IEEE 2022)

This special session, which is hosted in the IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2022) of WCCI 2022, will focus on AI-based systems relying on fuzzy neural networks (FNNs) and randomization principles, exploring the possibility of theorizing new potentials of interpretable FNNs. The application of layer randomization and FNNs techniques in energy-aware AI developments demonstrates the usefulness of such a learning approach in several practical applications, where explainable fuzzy models associated with fast and reliable training can give an effective contribution.