Performance measures for sparse spike inversion vs. basis pursuit inversion

Sparse Spike Inversion (SSI) and Basis Pursuit Inversion (BPI) are two methods for seismic inversion which utilize sparse inversion techniques. A LARS-LASSO solver is used for comparison of their performances. We propose the use of the F-measure to better evaluate the methods' capability to correctly identify reflections from layers' boundaries. We also show that flexible l1 penalization in the LASSO solution holds potential for improving performance.