Sound source ranging using a feed-forward neural network trained with fitting-based early stopping
Abstract
When a feed-forward neural network (FNN) is trained for source ranging in an ocean waveguide, it is difficult evaluating the range accuracy of the FNN on unlabeled test data. A fitting-based early stopping (FEAST) method is introduced to evaluate the range error of the FNN on test data where the distance of source is unknown. Based on FEAST, when the evaluated range error of the FNN reaches the minimum on test data, stopping training, which will help to improve the ranging accuracy of the FNN on the test data. The FEAST is demonstrated on simulated and experimental data.
- Publication:
-
Acoustical Society of America Journal
- Pub Date:
- September 2019
- DOI:
- 10.1121/1.5126115
- arXiv:
- arXiv:1904.00583
- Bibcode:
- 2019ASAJ..146L.258C
- Keywords:
-
- Computer Science - Machine Learning;
- Computer Science - Sound;
- Electrical Engineering and Systems Science - Signal Processing;
- Physics - Computational Physics;
- Statistics - Machine Learning
- E-Print:
- doi:10.1121/1.5126115