Published in Quantum Machine Intelligence, 2024
Collaboration between QuaTI and NTT Data. We applied variational circuits as confidence calibrators to classical language models for question-answering tasks, showing good improvements over the baseline.
Recommended citation: Jogi Suda Neto, Lluis Quiles Ardila, Thiago Nascimento Nogueira, Felipe Albuquerque, João Paulo Papa, Rodrigo Capobianco Guido & Felipe Fernandes Fanchini
Published in Elsevier Digital Signal Processing, 2021
Joint work with the Signal Processing lab at UNESP to develop a novel transform, designed to serve as a new tool for handcrafted feature extraction (FE). It improves the elementary analysis provided by signal energy (E) and enhances the humble spectral investigation granted by zero-crossing rates (ZCR). Python code is presented here for both the DPT as well as the inverse transform, IDPT, allowing for perfect reconstruction of the signal.
Recommended citation: Rodrigo Capobianco Guido, Fernando Pedroso, Rodrigo Colnago Contreras, Luciene Cavalcanti Rodrigues, Emanuel Guariglia, Jogi Suda Neto
Published in Elsevier Biomedical Signal Processing and Control, 2021
Undergraduate research dissertation at UNESP. A novel index for diagnosis of Keratoconus was proposed, which we called Corneal Tomography Multivariate Index (CTMVI). It uses Paraconsistent Feature Engineering and Support-Vector Machines to differentiate between healthy and pathological (clinical/subclinical) patients.
Recommended citation: Gildasio Castello de Almeida Jr, Rodrigo Capobianco Guido, Jogi Suda Neto, João Marcos Rosa, Lilian Castiglioni, Luiz Carlos de Mattos, Cinara Cássia Brandão.
Published in International Journal of Wavelets Multiresolution and Information Processing, 2020
Informative paper, joint collaboration with Signal Processing lab at UNESP.
Recommended citation: Guido, R. C., Pedroso, F., Furlan, A., Contreras, R. C., Caobianco, L. G., & Neto, J. S. (2020)