Abstract: The present invention relates to a method for calibrating on-line a direct neural interface implementing a REW-NPLS regression between an output calibration tensor and an input calibration tensor. The REW-NPLS regression comprises a PARAFAC iterative decomposition of the cross covariance tensor between the input calibration tensor and the output calibration tensor, each PARAFAC iteration comprising a sequence of M elementary steps (2401, 2401, . . . 240M) of minimisation of a metric according to the alternating least squares method, each elementary minimisation step relating to a projector and considering the others as constant, said metric comprising a penalisation term that is a function of the norm of this projector, the elements of this projector not being subjected to a penalisation during a PARAFAC iteration f not being penalisable during following PARAFAC iterations. Said calibration method makes it possible to obtain a predictive model of which the non-zero coefficients are sparse blockwise.

Abstract: The present invention relates to a method for calibrating on-line a direct neural interface implementing a REW-NPLS regression between an output calibration tensor and an input calibration tensor. The REW-NPLS regression comprises a PARAFAC iterative decomposition of the cross covariance tensor between the input calibration tensor and the output calibration tensor, each PARAFAC iteration comprising a sequence of M elementary steps (2401, 2401, . . . 240M) of minimisation of a metric according to the alternating least squares method, each elementary minimisation step relating to a projector and considering the others as constant, said metric comprising a penalisation term that is a function of the norm of this projector, the elements of this projector not being subjected to a penalisation during a PARAFAC iteration f not being penalisable during following PARAFAC iterations. Said calibration method makes it possible to obtain a predictive model of which the non-zero coefficients are sparse blockwise.

Abstract: This invention relates to a method of calibrating a direct neural interface with continuous coding. The observation variable is modelled by an HMM model and the control variable is estimated by means of a Markov mixture of experts, each expert being associated with a state of the model. During each calibration phase, the predictive model of each of the experts is trained on a sub-sequence of observation instants corresponding to the state with which it is associated, using an REW-NPLS (Recursive Exponentially Weighted N-way Partial Least Squares) regression model. A second predictive model giving the probability of occupancy of each state of the HMM model is also trained during each calibration phase using an REW-NPLS regression method. This second predictive model is used to calculate Markov mixture coefficients during a later operational prediction phase.