Optimizing ELARS Algorithms Using NVIDIA CUDA Heterogeneous Parallel Programming Platform