Journal: bioRxiv
Article Title: Evaluating Evolutionary and Gradient-Based Algorithms for Optimal Pathfinding
doi: 10.1101/2025.03.16.643541
Figure Lengend Snippet: The environment to traverse ( A ) and the simulation pipeline ( B ). The environment was generated as a linear combination of scaled and translated Gaussian surfaces, as described in Methods and Materials. The simulation procedure entailed refining the path predictions by each algorithm separately using the landscape-dependent cost function. Abbreviations: GA—genetic Algorithm; PSO—Particle Swarm Optimization; SQP— Sequential Quadratic Programming; a.u.—arbitrary units.
Article Snippet: The third algorithm used in our study was a quasi-Newton method—the Sequential Quadratic Programming (SQP) procedure (“ fmincon ” function in the MATLAB’s Optimization Toolbox)—described in detail in ( ).
Techniques: Generated, Refining