activity
20212024
most citedContinual-Learning-as-a-Service (CLaaS): On-Demand Efficient Adaptation of Predictive Models

4 citations · 10 across the 15 of their papers we have counts for

collaborators
Showing cs.LGShow all

12 papers · 1 filter

cs.LG2024

MultiSTOP: Solving Functional Equations with Reinforcement Learning

Alessandro Trenta, Davide Bacciu, Andrea Cossu +1

We develop MultiSTOP, a Reinforcement Learning framework for solving functional equations in physics. This new methodology produces actual numerical solutions instead of bounds on…

cs.LG2024

Calibration of Continual Learning Models

Lanpei Li, Elia Piccoli, Andrea Cossu +2

Continual Learning (CL) focuses on maximizing the predictive performance of a model across a non-stationary stream of data. Unfortunately, CL models tend to forget previous knowled…

cs.LG2024

Multi-Relational Graph Neural Network for Out-of-Domain Link Prediction

Asma Sattar, Georgios Deligiorgis, Marco Trincavelli +1

Dynamic multi-relational graphs are an expressive relational representation for data enclosing entities and relations of different types, and where relationships are allowed to var…

cs.LG20232 cited

Modeling Edge Features with Deep Bayesian Graph Networks

Daniele Atzeni, Federico Errica, Davide Bacciu +1

We propose an extension of the Contextual Graph Markov Model, a deep and probabilistic machine learning model for graphs, to model the distribution of edge features. Our approach i…

cs.LG2023

ADLER -- An efficient Hessian-based strategy for adaptive learning rate

Dario Balboni, Davide Bacciu

We derive a sound positive semi-definite approximation of the Hessian of deep models for which Hessian-vector products are easily computable. This enables us to provide an adaptive…

cs.LG20231 cited

Projected Latent Distillation for Data-Agnostic Consolidation in Distributed Continual Learning

Antonio Carta, Andrea Cossu, Vincenzo Lomonaco +2

Distributed learning on the edge often comprises self-centered devices (SCD) which learn local tasks independently and are unwilling to contribute to the performance of other SDCs.…