Cuckoo search with Lévy flights for weighted Bayesian energy functional optimization in global-support curve data...

Area of Science:

• Computational geometry
• Optimization algorithms
• Applied mathematics

Background:

• Data fitting is crucial in various scientific and engineering domains.
• Traditional methods often rely on piecewise functions (e.g., B-splines, NURBS) offering local control.
• Global-support curves, based on basis functions spanning the entire domain, present an alternative approach.

Purpose of the Study:

• To optimize a weighted Bayesian energy functional for data fitting.
• To explore the application of global-support approximating curves for this optimization problem.
• To evaluate the efficacy of the cuckoo search metaheuristic algorithm in this context.

Main Methods:

• Utilizing global-support curves, defined as linear combinations of basis functions with full domain support.
• Applying the cuckoo search algorithm, a nature-inspired metaheuristic known for its simplicity and few parameters.
• Testing the approach on diverse 2D and 3D curve fitting examples, including those with cusps and self-intersections.

Main Results:

• The cuckoo search algorithm effectively optimizes the weighted Bayesian energy functional for data fitting.
• The proposed method demonstrates successful application to various challenging 2D and 3D curve fitting scenarios.
• The simplicity of the cuckoo search algorithm facilitates straightforward parameter tuning.

Conclusions:

• Global-support approximating curves combined with cuckoo search offer a powerful and straightforward solution for data fitting optimization.
• The method's ability to handle complex curve features highlights its robustness.
• This approach presents a promising alternative to traditional data fitting techniques.