Abstract
As a tool for predicting stationary signals, the Least Mean Squares (LMS) algorithm is widely used. Its improvement, the family of normalised LMS algorithms, is known to outperform this algorithm. However, they still remain sensitive to selecting wrong parameters, being the learning coefficient μ and the signal history length M. We propose an improved version of both algorithms using a Fast Lane Approach, based on parallel evaluation of several competing predictors. These were applied to respiratory motion data from motion-compensated radiosurgery. Prediction was performed using arbitrarily selected values for the learning coefficient μ ∈] 0, 0.3] and the signal history length M ∈ [1, 15]. The results were compared to prediction using the globally optimal values of μ and M found using a grid search. When the learning algorithm is seeded using locally optimal values (found using a grid search on the first 96 s of data), more than 44% of the test cases outperform the globally optimal result. In about 38% of the cases, the result comes to within 5% and, in about 9% of the cases, to within 5-10% of the global optimum. This indicates that the Fast Lane Approach is a robust method for selecting the parameters μ and M.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 291-299 |
| Number of pages | 9 |
| Journal | Biomedical Signal Processing and Control |
| Volume | 3 |
| Issue number | 4 |
| DOIs | |
| State | Published - Oct 2008 |
| Externally published | Yes |
Fingerprint
Keywords
- LMS
- Parameter selection
- Radiosurgery
- Respiratory motion prediction
ASJC Scopus subject areas
- Health Informatics
- Signal Processing
Cite this
A Fast Lane Approach to LMS prediction of respiratory motion signals. / Ernst, Floris; Schlaefer, Alexander; Dieterich, Sonja; Schweikard, Achim.
In: Biomedical Signal Processing and Control, Vol. 3, No. 4, 10.2008, p. 291-299.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - A Fast Lane Approach to LMS prediction of respiratory motion signals
AU - Ernst, Floris
AU - Schlaefer, Alexander
AU - Dieterich, Sonja
AU - Schweikard, Achim
PY - 2008/10
Y1 - 2008/10
N2 - As a tool for predicting stationary signals, the Least Mean Squares (LMS) algorithm is widely used. Its improvement, the family of normalised LMS algorithms, is known to outperform this algorithm. However, they still remain sensitive to selecting wrong parameters, being the learning coefficient μ and the signal history length M. We propose an improved version of both algorithms using a Fast Lane Approach, based on parallel evaluation of several competing predictors. These were applied to respiratory motion data from motion-compensated radiosurgery. Prediction was performed using arbitrarily selected values for the learning coefficient μ ∈] 0, 0.3] and the signal history length M ∈ [1, 15]. The results were compared to prediction using the globally optimal values of μ and M found using a grid search. When the learning algorithm is seeded using locally optimal values (found using a grid search on the first 96 s of data), more than 44% of the test cases outperform the globally optimal result. In about 38% of the cases, the result comes to within 5% and, in about 9% of the cases, to within 5-10% of the global optimum. This indicates that the Fast Lane Approach is a robust method for selecting the parameters μ and M.
AB - As a tool for predicting stationary signals, the Least Mean Squares (LMS) algorithm is widely used. Its improvement, the family of normalised LMS algorithms, is known to outperform this algorithm. However, they still remain sensitive to selecting wrong parameters, being the learning coefficient μ and the signal history length M. We propose an improved version of both algorithms using a Fast Lane Approach, based on parallel evaluation of several competing predictors. These were applied to respiratory motion data from motion-compensated radiosurgery. Prediction was performed using arbitrarily selected values for the learning coefficient μ ∈] 0, 0.3] and the signal history length M ∈ [1, 15]. The results were compared to prediction using the globally optimal values of μ and M found using a grid search. When the learning algorithm is seeded using locally optimal values (found using a grid search on the first 96 s of data), more than 44% of the test cases outperform the globally optimal result. In about 38% of the cases, the result comes to within 5% and, in about 9% of the cases, to within 5-10% of the global optimum. This indicates that the Fast Lane Approach is a robust method for selecting the parameters μ and M.
KW - LMS
KW - Parameter selection
KW - Radiosurgery
KW - Respiratory motion prediction
UR - http://www.scopus.com/inward/record.url?scp=50249185259&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=50249185259&partnerID=8YFLogxK
U2 - 10.1016/j.bspc.2008.06.001
DO - 10.1016/j.bspc.2008.06.001
M3 - Article
AN - SCOPUS:50249185259
VL - 3
SP - 291
EP - 299
JO - Biomedical Signal Processing and Control
JF - Biomedical Signal Processing and Control
SN - 1746-8094
IS - 4
ER -