Daniele CASTELLANA

Ruolo attuale:: Ricercatore Legge 240/10 a tempo determinato
SSD:: INFO-01/A - Informatica
Afferenza organizzativa:: Dipartimento di Statistica, Informatica, Applicazioni 'G. Parenti' (DiSIA)

Recapiti

daniele.castellana(AT)unifi.it

Ulteriori Recapiti

https://danielecastellana22.github.io/

Daniele CASTELLANA

Orario di ricevimento (aggiornato al 06/09/2022)

Su appuntamento da fissare tramite e-mail nei giorni
Giovedì: 14:30 - 16:30
Venerdì: 11:00 - 13:00

Daniele CASTELLANA

Pubblicazioni

Legenda

Contributo su rivista |

Articolo su libro |

Libro |

Contributo in atti di convegno (proceeding) |

Brevetto |

Curatela | Altro

Altro |

Tesi di Dottorato

antonio carta; daniele castellana (2024). CD-IMM: The Benefits of Domain-based Mixture Models in Bayesian Continual Learning. In: Continual Artificial Intelligence Unconference, Neil Lawrence, pp. 0-0.

Daniele Castellana (2024). Lying Graph Convolution: Learning to Lie for Node Classification Tasks. In: International Joint Conference on Neural Networks (IJCNN), IEEE, pp. 0-0. DOI

Castellana Daniele, Errica Federico, Bacciu Davide, Micheli Alessio (2022). The Infinite Contextual Graph Markov Model. In: Internation Conference on Machine Learning, PMLR, vol. 162, pp. 2721-2737. Accesso ONLINE all'editore

Castellana D.; Bacciu D. (2022). A tensor framework for learning in structured domains. NEUROCOMPUTING, vol. 470, pp. 405-426, ISSN:0925-2312 DOI

Daniele Castellana (2021). A Tensor Framework for Learning in Structure Domains. (Tutor: Davide Bacciu)

Castellana, Daniele; Bacciu, Davide (2020). Learning from Non-Binary Constituency Trees via Tensor Decomposition. In: 28th International Conference on Computational Linguistics, International Committee on Computational Linguistics, pp. 3899-3910. DOI

Castellana D.; Bacciu D. (2020). Generalising Recursive Neural Models by Tensor Decomposition. In: 2020 International Joint Conference on Neural Networks, IJCNN 2020, gbr, 2020, Institute of Electrical and Electronics Engineers Inc., pp. 1-8, ISBN:978-1-7281-6926-2 DOI

Castellana D.; Bacciu D. (2020). Tensor decompositions in recursive neural networks for tree-structured data. In: 28th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2020, bel, 2020, ESANN, pp. 451-456.

Castellana D.; Bacciu D. (2019). Bayesian Tensor Factorisation for Bottom-up Hidden Tree Markov Models. In: 2019 International Joint Conference on Neural Networks, IJCNN 2019, hun, 2019, Institute of Electrical and Electronics Engineers Inc., vol. 2019-, pp. 1-8, ISBN:978-1-7281-1985-4 DOI

Davide Bacciu; Daniele Castellana (2019). Bayesian Mixtures of Hidden Tree Markov Models for Structured Data Clustering. NEUROCOMPUTING, vol. 342, pp. 49-59, ISSN:0925-2312 DOI

Davide Bacciu; CASTELLANA, DANIELE (2018). Mixture of Hidden Markov Models as Tree Encoder. In: European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN'18), Bruges, Belgium, 25-27 Aprile 2018, i6doc.com, pp. 543-548.

Davide Bacciu; CASTELLANA, DANIELE (2018). Learning Tree Distributions by Hidden Markov Models. In: Workshop on Learning and Automata (LearnAut'18), Oxford, UK, -, pp. 0-0. Accesso ONLINE all'editore

Daniele CASTELLANA

Biography

I am a Research Fellow (RTD-A) at the University of Florence in the DISIA deperatment. My interests are mainly related to Machine Learning, with emphasis on tensor decompositions in structured domains.

I obtained my PhD at the Universtiy of Pisa in 2021 under the supervision of Prof. Davide Bacciu. In Pisa, I was a member of the Computational Intelligence and Machine Learning group and Pervasive AI lab.

During the MSc, I had the opportunity to spend one year at the University College of London (UCL) as an Erasmus+ student. There, I attended Machine Learning MSc courses.

I also like to participate in Competitive Programming contests. In 2011, I won a silver medal at the OII, an italian competitive programming contest for high school students.

Daniele CASTELLANA

Curriculum

Research Experience

08/2021 – 08/2022	PostDoc on tensor-based ML approaches for strucutred data, Univerisità di Pisa, Pisa, Italia.
02/2021 – 07/2021	Scholarships on tensor decompositions in learning for structured data, Univerisità di Pisa, Pisa, Italia.

Education

2017–2021	Phd in Computer Science, Univerisità degli Studi di Pisa, Pisa, Italy, Ottimo con Lode. Thesis Title: "A Tensor Framework for Learning in Structured Domains", Supervisor: Prof. Davide Bacciu
2014–2017	Master’s Degree in Computer Science, Univerisità degli Studi di Pisa, Pisa, Italy, 110/110 with honours. Thesis Title: "Learning Tree Transducers: a coupled Hidden Markov Tree approach", Supervisor: Prof. Davide Bacciu
2015–2016	Msc in Machine Learning, Univeristy College of London, London, United Kingdom, 79.75%. Erasmus+ Affiliate Student
2011–2014	Bachelor’s Degree in Computer Science, Univeristà degli Studi di Pisa, Pisa, Italy, 110/110 with honours. Thesis Title: "Novel interface for novel input device", Supervisor: Prof. Antonio Cisterino

Daniele CASTELLANA

Interest

My research interests are related to Machine Learning, both from a theoretical and an application point of view.