Export Citations
Publications
Save this search
Please login to be able to save your searches and receive alerts for new content matching your search criteria.
- research-article
Hidden classification layers: Enhancing linear separability between classes in neural networks layers
Pattern Recognition Letters, Volume 177, Issue C•Jan 2024, pp 69-74 • https://doi.org/10.1016/j.patrec.2023.11.016AbstractMany Deep Learning approaches are based on variations of standard multi-layer feed-forward neural networks. These are also referred to as deep networks. The basic idea is that each hidden neural layer accomplishes a data transformation which is ...
Highlights- A neural network architecture inducing an error function involving the outputs of all the network layers is proposed.
- A high degree of linear separability between the classes in the hidden layers is induced by a proper loss function.
- research-article
Adaptive filters in Graph Convolutional Neural Networks
AbstractOver the last few years, the availability of an increasing data generated from non-Euclidean domains, which are usually represented as graphs with complex relationships, and Graph Neural Networks (GNN) have gained a high interest because of their ...
Highlights- The Convolutional Graph Neural Network parameters are set by another neural network.
- Filters are input-specific and dynamically generated exploiting the features of the graph nodes.
- The convolution is performed on each graph node ...
- Article
Dynamic Local Filters in Graph Convolutional Neural Networks
Image Analysis and Processing – ICIAP 2023•September 2023, pp 406-417• https://doi.org/10.1007/978-3-031-43153-1_34AbstractOver the last few years, we have seen increasing data generated from non-Euclidean domains, usually represented as graphs with complex relationships. Graph Neural Networks (GNN) have gained a high interest because of their potential in processing ...
- research-article
On the effects of data normalization for domain adaptation on EEG data
Engineering Applications of Artificial Intelligence, Volume 123, Issue PA•Aug 2023 • https://doi.org/10.1016/j.engappai.2023.106205AbstractIn Machine Learning (ML), a well-known problem is the Dataset Shift problem where the data in the training and test sets can follow different probability distributions, leading ML systems toward poor generalization performances. This problem is ...
Highlights- We study the impact of data normalization on DA for EEG classification problems.
- To the best of our knowledge, this aspect has yet to be extensively investigated.
- We show that normalization plays a key role in performance when ...
- research-article
Exploiting auto-encoders and segmentation methods for middle-level explanations of image classification systems
Knowledge-Based Systems, Volume 255, Issue C•Nov 2022 • https://doi.org/10.1016/j.knosys.2022.109725AbstractA central issue addressed by the rapidly growing research area of eXplainable Artificial Intelligence (XAI) is to provide methods to give explanations for the behaviours of Machine Learning (ML) non-interpretable models after the ...
- research-article
Adoption of Machine Learning Techniques to Enhance Classification Performance in Reactive Brain-Computer Interfaces
- Andrea Apicella,
- Pasquale Arpaia,
- Andrea Cataldo,
- Egidio De Benedetto,
- Nicola Donato,
- Luigi Duraccio,
- Salvatore Giugliano,
- Roberto Prevete
2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)•June 2022, pp 1-5• https://doi.org/10.1109/MeMeA54994.2022.9856441This paper proposes the adoption of an innovative algorithm to enhance the performance of highly wearable, reactive Brain-Computer Interfaces (BCIs), which exploit the Steady-State Visually Evoked Potential (SSVEP) paradigm. In particular, a combined time-...
- research-article
Neural Network-Based Prediction and Monitoring of Blood Glucose Response to Nutritional Factors in Type-1 Diabetes
- Leopoldo Angrisani,
- Giovanni Annuzzi,
- Pasquale Arpaia,
- Lutgarda Bozzetto,
- Andrea Cataldo,
- Alessandra Corrado,
- Egidio De Benedetto,
- Vincenzo Di Capua,
- Roberto Prevete,
- Ersilia Vallefuoco
2022 IEEE International Instrumentation and Measurement Technology Conference (I2MTC)•May 2022, pp 1-6• https://doi.org/10.1109/I2MTC48687.2022.9806611Type 1 diabetes (T1D) is an autoimmune disease that affects millions of people worldwide. A most challenging aspect regarding diabetes therapy is the way to calculate the insulin bolus amount to inject before meals. The artificial pancreas (AP), combining ...
- research-article
A ML-based Approach to Enhance Metrological Performance of Wearable Brain-Computer Interfaces
- Leopoldo Angrisani,
- Andrea Apicella,
- Pasquale Arpaia,
- Egidio De Benedetto,
- Nicola Donato,
- Luigi Duraccio,
- Salvatore Giugliano,
- Roberto Prevete
2022 IEEE International Instrumentation and Measurement Technology Conference (I2MTC)•May 2022, pp 1-5• https://doi.org/10.1109/I2MTC48687.2022.9806518In this paper, the adoption of Machine Learning (ML) classifiers is addressed to improve the performance of highly wearable, single-channel instrumentation for Brain-Computer Interfaces (BCIs). The proposed BCI is based on the classification of Steady-...
- Article
A General Approach to Compute the Relevance of Middle-Level Input Features
Pattern Recognition. ICPR International Workshops and Challenges•January 2021, pp 189-203• https://doi.org/10.1007/978-3-030-68796-0_14AbstractThis work proposes a novel general framework, in the context of eXplainable Artificial Intelligence (XAI), to construct explanations for the behaviour of Machine Learning (ML) models in terms of middle-level features which represent perceptually ...
- research-article
Preliminary validation of a measurement system for emotion recognition
2020 IEEE International Symposium on Medical Measurements and Applications (MeMeA)•June 2020, pp 1-6• https://doi.org/10.1109/MeMeA49120.2020.9137353An highly-wearable (wireless, few–channels and dry electrodes) device is proposed for EEG based valence emotion recognition. The component is a part of an instrument for real time engagement assessment in rehabilitation 4.0. The frontal, central, ...
- research-article
A simple and efficient architecture for trainable activation functions
Neurocomputing, Volume 370, Issue C•Dec 2019, pp 1-15 • https://doi.org/10.1016/j.neucom.2019.08.065AbstractAutomatically learning the best activation function for the task is an active topic in neural network research. At the moment, despite promising results, it is still challenging to determine a method for learning an activation function ...
- Article
Contrastive Explanations to Classification Systems Using Sparse Dictionaries
Image Analysis and Processing – ICIAP 2019•September 2019, pp 207-218• https://doi.org/10.1007/978-3-030-30642-7_19AbstractProviding algorithmic explanations for the decisions of machine learning systems to end users, data protection officers, and other stakeholders in the design, production, commercialisation and use of machine learning systems pipeline is an ...
- research-article
Learning programs is better than learning dynamics
Adaptive Behavior - Animals, Animats, Software Agents, Robots, Adaptive Systems, Volume 24, Issue 1•2 2016, pp 27-51 • https://doi.org/10.1177/1059712315609412Distributed and hierarchical models of control are nowadays popular in computational modeling and robotics. In the artificial neural network literature, complex behaviors can be produced by composing elementary building blocks or motor primitives, ...
- research-article
Neural networks with non-uniform embedding and explicit validation phase to assess Granger causality
- Alessandro Montalto,
- Sebastiano Stramaglia,
- Luca Faes,
- Giovanni Tessitore,
- Roberto Prevete,
- Daniele Marinazzo
Neural Networks, Volume 71, Issue C•November 2015, pp 159-171 • https://doi.org/10.1016/j.neunet.2015.08.003A challenging problem when studying a dynamical system is to find the interdependencies among its individual components. Several algorithms have been proposed to detect directed dynamical influences between time series. Two of the most used approaches ...
- article
Programming in the brain: a neural network theoretical framework
Recent research shows that some brain areas perform more than one task and the switching times between them are incompatible with learning and that parts of the brain are controlled by other parts of the brain, or are “recycled”, or are used and reused ...
- article
Perceiving affordances: A computational investigation of grasping affordances
Cognitive Systems Research, Volume 12, Issue 2•June, 2011, pp 122-133 • https://doi.org/10.1016/j.cogsys.2010.07.005The Grasping Affordance Model (GAM) introduced here provides a computational account of perceptual processes enabling one to identify grasping action possibilities from visual scenes. GAM identifies the core of affordance perception with visuo-motor ...
- Article
A robotic scenario for programmable fixed-weight neural networks exhibiting multiple behaviors
ICANNGA'11: Proceedings of the 10th international conference on Adaptive and natural computing algorithms - Volume Part I•April 2011, pp 250-259Artificial neural network architectures are systems which usually exhibit a unique/special behavior on the basis of a fixed structure expressed in terms of parameters computed by a training phase. In contrast with this approach, we present a robotic ...
- article
From motor to sensory processing in mirror neuron computational modelling
Biological Cybernetics, Volume 103, Issue 6•December 2010, pp 471-485 • https://doi.org/10.1007/s00422-010-0415-5Typical patterns of hand-joint covariation arising in the context of grasping actions enable one to provide simplified descriptions of these actions in terms of small sets of hand-joint parameters. The computational model of mirror mechanisms introduced ...
- Article
A Semi-automated Method for the Measurement of the Fetal Nuchal Translucency in Ultrasound Images
ICIAP '09: Proceedings of the 15th International Conference on Image Analysis and Processing•August 2009, pp 613-622• https://doi.org/10.1007/978-3-642-04146-4_66Nowadays the measurement of the nuchal translucency thickness is being used as part of routine ultrasound scanning during the end of the first trimester of pregnancy, for the screening of chromosomal defects, as trisomy 21. Currently, the measurement is ...
- Article
CTRNN Parameter Learning using Differential Evolution
- Ivanoe De Falco,
- Antonio Della Cioppa,
- Francesco Donnarumma,
- Domenico Maisto,
- Roberto Prevete,
- Ernesto Tarantino
Proceedings of the 2008 conference on ECAI 2008: 18th European Conference on Artificial Intelligence•June 2008, pp 783-784Target behaviours can be achieved by finding suitable parameters for Continuous Time Recurrent Neural Networks (CTRNNs) used as agent control systems. Differential Evolution (DE) has been deployed to search parameter space of CTRNNs and overcome ...