Abstract
Brain Machine Interfaces (BMI) utilize linear or non-linear models to map the neural activity to the associated behavior which is typically the 2-D or 3-D hand position of a primate. Linear models are plagued by the massive disparity of the input and output dimensions thereby leading to poor generalization. A solution would be to use non-linear models like the Recurrent Multi-Layer Perceptron (RMLP) that provide parsimonious mapping functions with better generalization. However, this results in a drastic increase in the training complexity, which can be critical for practical use of a BMI. This paper bridges the gap between superior performance per trained weight and model learning complexity. Towards this end, we propose to use Echo State Networks (ESN) to transform the neuronal firing activity into a higher dimensional space and then derive an optimal sparse linear mapping in the transformed space to match the hand position. The sparse mapping is obtained using a weight constrained cost function whose optimal solution is determined using a stochastic gradient algorithm.
| Original language | English |
|---|---|
| Title of host publication | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Volume | V |
| ISBN (Print) | 0780388747, 9780780388741 |
| DOIs | |
| State | Published - Jan 1 2005 |
| Externally published | Yes |
| Event | 2005 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP '05 - Philadelphia, PA, United States Duration: Mar 18 2005 → Mar 23 2005 |
Other
| Other | 2005 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP '05 |
|---|---|
| Country | United States |
| City | Philadelphia, PA |
| Period | 3/18/05 → 3/23/05 |
Fingerprint
ASJC Scopus subject areas
- Acoustics and Ultrasonics
- Electrical and Electronic Engineering
- Signal Processing
Cite this
Learning mappings in brain machine interfaces with echo state networks. / Rao, Yadunandana N.; Kim, Sung Phil; Sanchez, Justin C.; Erdogmus, Deniz; Principe, Jose C.; Carmena, Jose M.; Lebedev, Mikhail A.; Nicolelis, Miguel A.
ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. Vol. V Institute of Electrical and Electronics Engineers Inc., 2005. 1416283.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Learning mappings in brain machine interfaces with echo state networks
AU - Rao, Yadunandana N.
AU - Kim, Sung Phil
AU - Sanchez, Justin C.
AU - Erdogmus, Deniz
AU - Principe, Jose C.
AU - Carmena, Jose M.
AU - Lebedev, Mikhail A.
AU - Nicolelis, Miguel A.
PY - 2005/1/1
Y1 - 2005/1/1
N2 - Brain Machine Interfaces (BMI) utilize linear or non-linear models to map the neural activity to the associated behavior which is typically the 2-D or 3-D hand position of a primate. Linear models are plagued by the massive disparity of the input and output dimensions thereby leading to poor generalization. A solution would be to use non-linear models like the Recurrent Multi-Layer Perceptron (RMLP) that provide parsimonious mapping functions with better generalization. However, this results in a drastic increase in the training complexity, which can be critical for practical use of a BMI. This paper bridges the gap between superior performance per trained weight and model learning complexity. Towards this end, we propose to use Echo State Networks (ESN) to transform the neuronal firing activity into a higher dimensional space and then derive an optimal sparse linear mapping in the transformed space to match the hand position. The sparse mapping is obtained using a weight constrained cost function whose optimal solution is determined using a stochastic gradient algorithm.
AB - Brain Machine Interfaces (BMI) utilize linear or non-linear models to map the neural activity to the associated behavior which is typically the 2-D or 3-D hand position of a primate. Linear models are plagued by the massive disparity of the input and output dimensions thereby leading to poor generalization. A solution would be to use non-linear models like the Recurrent Multi-Layer Perceptron (RMLP) that provide parsimonious mapping functions with better generalization. However, this results in a drastic increase in the training complexity, which can be critical for practical use of a BMI. This paper bridges the gap between superior performance per trained weight and model learning complexity. Towards this end, we propose to use Echo State Networks (ESN) to transform the neuronal firing activity into a higher dimensional space and then derive an optimal sparse linear mapping in the transformed space to match the hand position. The sparse mapping is obtained using a weight constrained cost function whose optimal solution is determined using a stochastic gradient algorithm.
UR - http://www.scopus.com/inward/record.url?scp=33646777474&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33646777474&partnerID=8YFLogxK
U2 - 10.1109/ICASSP.2005.1416283
DO - 10.1109/ICASSP.2005.1416283
M3 - Conference contribution
AN - SCOPUS:33646777474
SN - 0780388747
SN - 9780780388741
VL - V
BT - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
ER -