FMM-Net: neural network architecture based on the Fast Multipole Method
This work addresses memory and efficiency issues for neural network practitioners, though it appears incremental as it builds on existing H2-based architectures.
The authors tackled the problem of high memory costs and performance limitations in neural networks by proposing FMM-Net, a new architecture based on the H2 matrix, which demonstrated benefits in performance, memory usage, and scalability compared to existing alternatives.
In this paper, we propose a new neural network architecture based on the H2 matrix. Even though networks with H2-inspired architecture already exist, and our approach is designed to reduce memory costs and improve performance by taking into account the sparsity template of the H2 matrix. In numerical comparison with alternative neural networks, including the known H2-based ones, our architecture showed itself as beneficial in terms of performance, memory, and scalability.