Machine Learning Group


|

Table of Contents

Brain-Computer Interfaces

Future Papers

  1. QDA with MSPRT
  2. QDA with MSPRT and feature selection
  3. Results from multiple users of our BCI using above
  4. Incremental QDA (Malai)

Current Papers

Software

MNF

  1. Implement MNF in R for artifact removal
  2. Add to R BCI code
  3. Display MNF components on head in R

P300 Speller

  1. Create R version for P300 speller
  2. OpenGL or GTK version of P300 speller
  3. Possible integration into CEBL
  4. Try using QDA and MNF
  5. Try template-based approaches
  6. Try using recurrent networks to model “typical” P300
  7. word prediction, T9 for example
  8. Here are some interesting links:

CEBL

  1. fix bugs in CEBL
  2. add Neural Network classifier module
  3. add SVM classifier module

Recording

  1. Continue collecting data
  2. Experiment with new mental tasks
  3. Investigate dry electrodes

BCI Applications

  1. Brain Maze