Manifesto for transparent mathematical modeling: from ecology to general science

Mathematical black-box models, which hide the structure and behavior of the subsystems, currently dominate science. Mechanisms under study remain hidden. Errors and paradoxes, such as the biodiversity paradox and the limiting similarity hypothesis, often arise from subjective interpretati...

Full description

Saved in:
Bibliographic Details
Main Authors: Vyacheslav L. Kalmykov, Lev V. Kalmykov
Format: Article
Language:English
Published: Academia.edu Journals 2024-01-01
Series:Academia Biology
Online Access:https://www.academia.edu/113709506/Manifesto_for_transparent_mathematical_modeling_from_ecology_to_general_science
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Mathematical black-box models, which hide the structure and behavior of the subsystems, currently dominate science. Mechanisms under study remain hidden. Errors and paradoxes, such as the biodiversity paradox and the limiting similarity hypothesis, often arise from subjective interpretations of these hidden mechanisms. To address these problems in ecology, we have developed transparent mathematical models of the white-box type. Here we present and justify the hypothesis that it is possible to construct transparent mathematical white-box models using logical deterministic cellular automata, where the rules used to construct these models are based on the general theory of the relevant domain. So far, white-box modeling has allowed us to directly identify the mechanisms of interspecific competition, test the principle of competitive exclusion and the hypothesis of limiting similarity, resolve the paradox of biodiversity, and formulate for the first time the general principle of competitive coexistence. As a framework for reproducing and further developing the method, we present a C++ code of two transparent mathematical models of an ecosystem. A shift to transparency in the mathematical modeling paradigm has the potential to revolutionize scientific research and to advance knowledge and technology in a wide variety of domains.
ISSN:2837-4010