Humberto Sandmann

Computer scientist and professor with a D.Sc. from the University of São Paulo and a research stay at the Max Planck Institute for Dynamics and Self-Organization. My research bridges bio-inspired neural computing, nonlinear dynamics, and pattern recognition, investigating how biological neuronal mechanisms can inform computational intelligence. With more than a decade of university teaching alongside senior industry practice in software architecture, artificial intelligence, and high-performance computing, I bring research depth and applied engineering experience into the classroom.

Research Interests

DeFi, smart contracts and decentralized applications
Small-world networks and complex systems
Computation in large-scale systems
Artificial intelligence and deep learning
Nonlinear systems and network dynamics
Synchronization of coupled oscillators

Education

mar.2012 ↑ mar.2007	D.Sc.Polytechnic School of the University of Sao Paulo thesisSpike patterns and computation in dynamical neural networks doi10.11606/T.3.2012.tde-05092012-165022 advisorProf. Dr. Emilio Del Moral Hernandez presentationslides supported by CNPq - grant 140632/2007-7 (mar.2007 → dec.2009) ∧ (sep.2010 → feb.2011)
sep.2010 ↑ aug.2009	D.Sc. InternshipMax Planck Gesellschaft für Dynamik und Selbstorganisation advisorProf. Dr. Marc Timme supported by CAPES - grant BEX 4076-09-5 jan.2010 → aug.2010 Pró-Reitoria da Pós-Graduação da USP aug.2009 ∧ dec.2009 (flight tickets GRU ⇄ FRA) José Humberto Sandmann - my father aug.2009 → dec.2009 (German language classes/exams) Max Planck Institute - Networks Dynamics Group aug.2009 → nov.2009 (accommodation in Goettingen)
jun.2006 ↑ feb.2003	M.Sc.Polytechnic School of the University of Sao Paulo dissertationPrediction of time series using architecture based on neuro-fuzzy systems doi10.11606/D.3.2006.tde-01042009-095125 advisorProf. Dr. Marco Túlio Carvalho de Andrade
dec.2002 ↑ feb.1999	B.Sc.Faculdade de Engenharia Industrial Computer Science Bachelor
dec.1998 ↑ feb.1996	Escola Técnica Jorge Street Technician in Industrial Computation

Academic Appointments & Teaching

Professional Experience

Publications

Selected peer-reviewed publications. Full and up-to-date list: ORCID 0000-0001-7758-695X.

H. Sandmann; L.A. Silva; E. Del-Moral-Hernandez. A self-organizing architecture of recursive elements for continuous learning (poster abstract). In Dynamics Days Europe 2009, Göttingen, Germany, 2009, pp. 246–248.

E. Del-Moral-Hernandez; H. Sandmann; G. Araujo. Context dependent pattern recognition — A framework for hybrid architectures bridging chaotic neural networks based on Recursive Processing Elements and symbolic information. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Atlanta, GA, USA, 2009, pp. 663–670. IEEE. doi: 10.1109/IJCNN.2009.5179061.

L.A. da Silva; H. Sandmann; E. Del-Moral-Hernandez. A self-organizing architecture of recursive elements for continuous learning. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Hong Kong, China, 2008 (IEEE World Congress on Computational Intelligence). IEEE. doi: 10.1109/IJCNN.2008.4634190.

H. Sandmann; M.T.C. de Andrade. Análise da capacidade de uma rede neural de memória associativa com arquitetura de elementos processados recursivamente. In Brazilian Symposium on Neural Networks (SBRN), São Luís, MA, Brazil, 2004. IEEE. pdf.

E. Del-Moral-Hernandez; H. Sandmann; L.A. da Silva. Pattern recovery in networks of recursive processing elements with continuous learning. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, 2004. IEEE. doi: 10.1109/IJCNN.2004.1379877.

Grants & Funding

Academic Service & Memberships

Conference Service

Program Committee member and reviewer for the International Joint Conference on Neural Networks (IJCNN) — the flagship conference of the INNS and the IEEE Computational Intelligence Society.

Media & Press

Interviews and expert commentary in Brazilian broadcast and print media.

Languages

This CV website was designed and built with the assistance of AI — Anthropic's Claude. All content is curated and verified by Humberto Sandmann.