Quasi-Dynamic System Representation of Multiple Nonlinear Regression

Explore a groundbreaking approach to multiple nonlinear regression analysis in this conference talk from the International Conference on Biological Sciences (ICBS) 2024. Delve into how evolutionary game theory and allometric scaling law can be combined to create a quasi-dynamic system representation that captures complex nonlinear relationships in nature. Learn about the innovative use of Legendre Orthogonal Polynomials to linearize nonlinear payoffs, enabling variable selection and addressing overfitting challenges. Discover how this novel method transforms regression results into bidirectional, signed, and weighted networks, providing a causal roadmap of variable interactions. Gain unique insights into revealing nonlinear causality across graded samples, offering a powerful new tool for researchers and data scientists working with complex biological systems.

Rongling Wu: A quasi-dynamic system representation of multiple nonlinear regression #ICBS2024