From Data to Combinatorial Multivector field Through an Optimization-Based Framework

Abstract: This paper extends and generalizes previous works on constructing combinatorial multivector fields from continuous systems (see [10]) and the construction of combinatorial vector fields from data (see [2]) by introducing an optimization based framework for the construction of combinatorial multivector fields from finite vector field data. We address key challenges in convexity, computational complexity and resolution, providing theoretical guarantees and practical methodologies for generating combinatorial representation of the dynamics of our data.