Monte Carlo Physarum Machine: Characteristics of Pattern Formation in Continuous Stochastic Transport Networks.

Area of Science:

• Computational modeling
• Astroparticle physics
• Network reconstruction

Background:

• Agent-based models simulate biological growth patterns.
• Reconstructing continuous transport networks from sparse data is challenging.
• Understanding the cosmic web requires mapping large-scale structures.

Purpose of the Study:

• Introduce the Monte Carlo Physarum Machine (MCPM) as a novel computational model.
• Generalize existing agent-based models for broader applications.
• Apply MCPM to reconstruct the cosmic web from cosmological data.

Main Methods:

• Developed MCPM as a probabilistic generalization of an agent-based slime mold growth model.
• Explored MCPM's self-patterning behavior to generate diverse network morphologies (polyphorms).
• Applied MCPM to simulated and observational cosmological data to create 3D density maps.

Main Results:

• MCPM generates a wide range of continuous network-like structures (polyphorms).
• The model successfully produces consistent 3D density maps of the cosmic web from cosmological data.
• Demonstrated MCPM's potential for various domain-specific data fitting tasks.

Conclusions:

• MCPM is a versatile tool for reconstructing continuous transport networks.
• The model offers a novel approach for analyzing cosmological structures like the cosmic web.
• MCPM has potential applications beyond cosmology, including fitting domain-specific data.