Geoacoustic Inversion Based on Support Vector Machine

This paper proposed an efficient inversion scheme based on support vector machine (SVM) to solve the geoacoustic inversion problem (GIP). The key of this inversion scheme is to use SVM to obtain the approximated inverse functions between the unknown geoacoustic parameters and the measured acoustic pressure amplitudes. The main advantage of this scheme is that the SVM-approximation-function above can be computed off-line from simulated data as soon as the experimental configuration is known. Then the solutions of GIP can be directly estimated using such approximation functions in true real time as soon as the data are acquired. In the end, the Mediterranean Sea experimental results validate our proposed inversion scheme.