General optimization framework for accurate and efficient reconstruction of symmetric complex networks from dynamical...

Area of Science:

• Complex Systems
• Network Science
• Computational Biology

Background:

• Network reconstruction from dynamical data is a critical challenge.
• Existing methods, point-by-point (PBP) and global, have trade-offs between computational efficiency and accuracy.
• Global methods offer high accuracy but are computationally expensive.

Purpose of the Study:

• To develop a novel network reconstruction framework that integrates the strengths of PBP and global methods.
• To achieve high reconstruction accuracy with significantly reduced computational cost.
• To leverage network symmetry for efficient inverse problem solving.

Main Methods:

• Developed a novel framework based on block coordinate descent (BCD).
• Utilized network symmetry to define blocks for BCD optimization.
• Validated the framework on diverse network structures (sparse and dense) and dynamical processes.

Main Results:

• The BCD-based framework achieves high network reconstruction accuracy.
• Computational cost is orders of magnitude lower than traditional global methods.
• Demonstrated effectiveness across various network types and dynamics.

Conclusions:

• Exploiting network symmetry is a powerful strategy for inverse problems.
• The BCD-based framework offers an efficient and accurate solution for network reconstruction.
• This approach has broad applicability in science and engineering problems involving linear equations.