MODAL research group at UNINA

We are thrilled to announce the publication of our research article:

"A Robust and Rapid Grid-Based Machine Learning Approach for Inside and Off-Network Earthquakes Classification in Dynamically Changing Seismic Networks" in Seismological Research Letters.

This groundbreaking work highlights a novel Machine Learning approach for improving earthquake classification, even in dynamically evolving seismic networks.

This achievement was made possible thanks to a fruitful collaboration between our research lab, Stanford University, and the National Institute of Geophysics and Volcanology (INGV).

Congratulations to the incredible team: Daniela Annunziata, Martina Savoia, Claudio Martino, Fabio Giampaolo, Vincenzo Convertito, Francesco Piccialli, and Gregory C. Beroza.

👉 Read more here: https://doi.org/10.1785/0220240173

We’re thrilled to share our new paper: "Enhancing De Novo Drug Design across Multiple Therapeutic Targets with CVAE Generative Models!"

Drug discovery is no easy task—it’s time-consuming and expensive. That’s why we explored how deep learning can revolutionize the process. Using a Conditional Variational Autoencoder (CVAE), we developed a generative model that creates new molecular designs tailored to specific therapeutic targets, validated for uniqueness, drug-likeness, synthetic accessibility, and more.

Our model showed great results for three key targets: CDK2, PPARγ, and DPP-IV, delivering diverse and promising molecules for drug discovery.

Our heartfelt thanks go to our amazing collaborators and the colleagues from the Department of Pharmacy, whose contributions highlight the importance of multidisciplinary studies in advancing research and innovation.

📖 Dive into the details: https://pubs.acs.org/doi/full/10.1021/acsomega.4c08027

Enhancing De Novo Drug Design across Multiple Therapeutic Targets with CVAE Generative Models Drug discovery is a costly and time-consuming process, necessitating innovative strategies to enhance efficiency across different stages, from initial hit identification to final market approval. Recent advancement in deep learning (DL), particularly in de novo drug design, show promise. Generative....

We’re excited to share our latest paper: "Federated and Edge Learning for Large Language Models"

With language models becoming more advanced, we looked at how to make them work better in federated and edge environments—balancing efficiency, privacy, and the challenges of limited resources. We explored cool techniques like model pruning and quantization to help LMs perform smarter at the edge.

A big shoutout to our amazing collaborators Pian Qi, Diletta Chiaro and colleagues from the University of Milan Valerio Bellandi Ernesto Damiani—your support made this possible! 💡

Prouder than ever to be part of this wonderful University of Naples Federico II!

📖 Read more: https://www.sciencedirect.com/science/article/pii/S1566253524006183

www.sciencedirect.com

🚀 CLAIM PROJECT: Empowering SMEs with AI-driven learning! 🌱🤖

The transition to a green and digital world is an extraordinary opportunity, but also a challenge in terms of skills. 🔨🌱
While the labour market requires people who can drive innovation and technological sustainability, many people find it difficult to upgrade.

This is where artificial intelligence comes in. An innovative platform powered by artificial intelligence is offered by Claim to help SMEs analyse and identify the training needs of their workforce. The platform pays special attention to crucial topics such as internationalisation and the digital and green transition. Claim enables companies to increase competitiveness and promote excellence in the changing global marketplace by analysing employee knowledge, skills and techniques.

An unprecedented challenge and opportunity is the integration of artificial intelligence into workforce training. Companies that manage to go through this transition will be able to take advantage of the enormous potential that artificial intelligence offers, while keeping their employees at the centre of this transformative change. It is crucial to invest in continuous training, acquire soft and technical skills and create a flexible and welcoming learning environment.

Exciting News! 📢

We are thrilled to announce that our Call for Papers for the FLBD 2024: Special Session on Federated Learning on Big Data has been published on !

Join us in exploring the cutting-edge advancements in Federated Learning and Big Data at the upcoming # IEEE Big Data 2024 conference. This special session aims to bring together researchers, practitioners, and industry experts to discuss innovative approaches, challenges, and future directions in this rapidly evolving field.

🗓 Submission Deadline: Sep 27, 2024
🌐 Learn More & Submit Your Paper: https://www3.cs.stonybrook.edu/~ieeebigdata2024/SpecialSessions.html

Don't miss this opportunity to showcase your research and contribute to the future of Federated Learning on Big Data. We look forward to your submissions!

FLBD 2024 : Special Session on Federated Learning on Big Data FLBD 2024 : Special Session on Federated Learning on Big Data

Federico II protagonista della Doctoral Network TUAI https://www.unina.it/-/60334564-federico-ii-protagonista-della-doctoral-network-tuai

Excited to share our poster presentation on the PRIN DIRECTIONS project at the INGV Rome conference! 🖼️

Our poster will highlight key insights and developments in seismic monitoring and early warning systems, emphasizing the innovative approaches and technologies employed to mitigate seismic risks effectively.

Join us on May 23rd as we showcase our research and findings.

🎉 We are incredibly excited to announce the very first special session "Federated Learning on Big Data" at the conference, which we are proudly organizing!

The session will highlight recent innovations in federated learning algorithms and frameworks designed to tackle the unique challenges posed by Big Data, such as heterogeneous data distributions and resource constraints.

So, what are you waiting for? Submit your paper and be a part of this exciting event! 📢

All the details are in the flyer and also @ https://www3.cs.stonybrook.edu/~ieeebigdata2024/SpecialSessions.html

See you in December in Washington! 🌟🗽

Mark your calendars for our Special Session and share this post to help spread the word! 🙌
🗓️ Stay tuned for more details and our session's exciting agenda.

Today we have joined Ital-IA 2024, the fourth National CINI Conference on Artificial Intelligence, organized to foster common goals among public institutions, the Italian industry, and the scientific research community from universities and national research centers. We were excited to present four exciting contributions regarding applications in , and ! 🇮🇹



https://ital-ia2024.it

We are pleased to announce that our paper, "FLOWS: Federated Learning Optimization with Sinkhorn," has been accepted at DistInSys @ 29th IEEE Symposium on Computers and Communications (ISCC 2024)! In this paper, we propose a new plug-in regularization solution to address the severe non-IID-ness of clients. 📄

We are excited to share our findings with this rich community!

FCRLab Future Research Computing Laboratory

We are thrilled to announce that our latest paper, "KAFÈ: Kernel Aggregation for FEderated Learning," has been accepted at , the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases! 🎉

KAFÈ addresses the non-IID problem in Federated Learning by leveraging Kernel Aggregation of the classification parameters on the server side. This is a significant milestone for our research team, as being accepted at ECML is a prestigious recognition of our work, and we are incredibly proud of this achievement.

Find out more here: https://ecmlpkdd.org/2024/program-accepted-papers-research-track/

See you in Vilnius in September!

ecmlpkdd.org