AutoRegressive Coefficients


  • Plugin name : AutoRegressive Coefficients
  • Version : 1.0
  • Author : Alison Cellard
  • Company : Inria
  • Short description : Estimates autoregressive (AR) coefficients from a set of signals
  • Documentation template generation date : Dec 30 2016
  • WARNING : this box has been marked as UNSTABLE by the developer. It means that its implementation may be incomplete or that the box can only work under well known conditions. It may possibly crash or cause data loss. Use this box at your own risk, you've been warned.


Estimates autoregressive (AR) linear model coefficients using Burg's method

The AR features box calculate the coefficients using Burg's method [1] to compute the AutoRegressive (AR) model of an input signal. The AR model is a representation that describes a time varying process by its own previous values.

The definition used is :


Where a(i) are the autoregressive coefficients or parameters of the model, x(t) is the input signal, x(t-i) its previous values, N is the order (length) of the model and epsilon(t) is the residue, assumed to be Gaussian white noise.

For more informations about AR model :

The model order (see [2]) needs to be specified in the settings of the box.

[1] Burg, J.P. (1967) "Maximum Entropy Spectral Analysis", Proceedings of the 37th Meeting of the Society of Exploration Geophysicists, Oklahoma City, Oklahoma

[2] D.J. Krusienski, D.J. MacFarland, J.R. Wolpaw. An evaluation of autoregressive spectral estimation model order for brain-computer interface application. Proceedings of the 28th IEEE EMBS Annual International Conference, New York City, USA, Aug 30-Sept 3, 2006


1. EEG Signal

The input signal

  • Type identifier : Signal (0x5ba36127, 0x195feae1)


1. AR Features

The AR coefficients stored in a Feature vector

  • Type identifier : Streamed matrix (0x544a003e, 0x6dcba5f6)


1. Order

Specify the order, thus the number of coefficients calculated

  • Type identifier : Integer (0x007deef9, 0x2f3e95c6)
  • Default value : [ 1 ]



The output feature vector contains the coefficients for each channel : the first [order+1] elements are the coefficients of the first channel, etc.