%0 Conference Proceedings %T Evidence build-up facilitates on-line adaptivity in dynamic environments: example of the BCI P300-speller %+ Institut de Neurosciences des Systèmes (INS) %+ Computational Imaging of the Central Nervous System (ATHENA) %+ Sequential Learning (SEQUEL) %A Daucé, Emmanuel %A Thomas, Eoin %< avec comité de lecture %B 22nd European Symposium on Artificial Neural Networks %C Bruges, Belgium %8 2014-04-23 %D 2014 %Z Statistics [stat]/Machine Learning [stat.ML] %Z Cognitive science/Computer science %Z Life Sciences [q-bio]/Bioengineering/ImagingConference papers %X We consider a P300 BCI application where the subjects can write figures and letters in an unsupervised fashion. We (i) show that a generic speller can attain the state-of-the-art accuracy without any training phase or calibration and (ii) present an adaptive setup that consistently increases the bit rate for most of the subjects. %G English %2 https://inria.hal.science/hal-01104024/document %2 https://inria.hal.science/hal-01104024/file/es2014-188.pdf %L hal-01104024 %U https://inria.hal.science/hal-01104024 %~ INSERM %~ UNIV-LILLE3 %~ CNRS %~ INRIA %~ UNIV-AMU %~ INRIA-SOPHIA %~ INRIA-LILLE %~ INRIASO %~ LAGIS %~ INRIA_TEST %~ TESTALAIN1 %~ INRIA2 %~ ANR %~ INS