Browse ORBilu by Open Access
- What it is and what it isn't
- Green Road / Gold Road?
- Ready to Publish. Now What?
- How can I support the OA movement?
- Where can I learn more?
ORBilu is a project developed by
  Consensual Resilient Control: (Don’t let the hackers spill your coffee) - Demo PosterMatovic, Aleksandar  ; Graczyk, Rafal ; Lucchetti, Federico et alPoster (2023, July 13) Detailed reference viewed: 56 (3 UL) Toward resilient autonomous driving—An experience report on integrating resilience mechanisms into the Apollo autonomous driving software stackLucchetti, Federico ; Graczyk, Rafal ; Volp, Marcus in Frontiers in Computer Science (2023), 5 Detailed reference viewed: 53 (0 UL)Tropical BackpropagationCeyhan, Özgür ; Lucchetti, Federico E-print/Working paper (2023) This work introduces tropicalization, a novel technique that delivers tropical neural networks as tropical limits of deep ReLU networks. Tropicalization transfers the initial weights from real numbers to ... [more ▼] This work introduces tropicalization, a novel technique that delivers tropical neural networks as tropical limits of deep ReLU networks. Tropicalization transfers the initial weights from real numbers to those in the tropical semiring while maintain- ing the underlying graph of the network. After verifying that tropicalization will not affect the classification capacity of deep neural networks, this study introduces a tropical reformulation of backpropagation via tropical linear algebra. Tropical arithmetic replaces multiplication operations in the network with additions and addition operations with max, and therefore, theoretically, reduces the algorithmic complexity during the training and inference phase. We demonstrate the latter by simulating the tensor multiplication underlying the feed-forward process of state- of-the-art trained neural network architectures and compare the standard forward pass of the models with the tropical ones. Our benchmark results show that tropi- calization speeds up inference by 50 %. Hence, we conclude that tropicalization bears the potential to reduce the training times of large neural networks drastically. [less ▲] Detailed reference viewed: 177 (1 UL)Federated Geometric Monte Carlo Clustering to Counter Non-IID DatasetsLucchetti, Federico ; ; et alE-print/Working paper (2022) Federated learning allows clients to collaboratively train models on datasets that are acquired in different locations and that cannot be exchanged because of their size or regulations. Such collected ... [more ▼] Federated learning allows clients to collaboratively train models on datasets that are acquired in different locations and that cannot be exchanged because of their size or regulations. Such collected data is increasingly non-independent and non- identically distributed (non-IID), negatively affecting training accuracy. Previous works tried to mitigate the effects of non- IID datasets on training accuracy, focusing mainly on non-IID labels, however practical datasets often also contain non-IID features. To address both non-IID labels and features, we propose FedGMCC1, a novel framework where a central server aggregates client models that it can cluster together. FedGMCC clustering relies on a Monte Carlo procedure that samples the output space of client models, infers their position in the weight space on a loss manifold and computes their geometric connection via an affine curve parametrization. FedGMCC aggregates connected models along their path connectivity to produce a richer global model, incorporating knowledge of all connected client models. FedGMCC outperforms FedAvg and FedProx in terms of convergence rates on the EMNIST62 and a genomic sequence classification datasets (by up to +63%). FedGMCC yields an improved accuracy (+4%) on the genomic dataset with respect to CFL, in high non-IID feature space settings and label incongruency. [less ▲] Detailed reference viewed: 78 (4 UL) Subcortical Neural Generators of the Envelope-Following Response in Sleeping Children: A Transfer Function AnalysisLucchetti, Federico in Hearing Research (2020) Detailed reference viewed: 118 (14 UL)Effect of vagus nerve stimulation on EEG oscillations and connectivity; ; Lucchetti, Federico et alin Effect of vagus nerve stimulation on EEG oscillations and connectivity (2020) Detailed reference viewed: 52 (0 UL)Automated epileptic seizure detection based on break of excitation/inhibition balance; ; et al in Computers in Biology and Medicine (2019) Detailed reference viewed: 74 (0 UL)Seizure evolution can be characterized as path through synaptic gain space of a neural mass model; ; et al in European Journal of Neuroscience (2018) Detailed reference viewed: 50 (0 UL)Generalization of the primary tone phase variation method: An exclusive way of isolating the frequency-following response componentsLucchetti, Federico ; ; et alin The Journal of the Acoustical Society of America (2018) Detailed reference viewed: 55 (0 UL)Dynamics underlying interictal to ictal transition in temporal lobe epilepsy: insights from a neural mass model; ; et al in European Journal of Neuroscience (2017) Detailed reference viewed: 75 (0 UL) |
||
