Handling uncertainty in portfolio optimization: A neutrosophic logic adaptive neural network solver for quadratic...

Area of Science:

• Robotics and Artificial Intelligence
• Computational Mathematics
• Optimization Theory

Background:

• Time-varying quadratic programming (TVQP) is crucial in fields like robotics and AI.
• Existing zeroing neural network (ZNN) techniques show promise for time-varying problems.
• There is a need for advanced solvers to handle the complexities of TVQP.

Purpose of the Study:

• Introduce a novel adaptive neutrosophic logic/fuzzy neural network TVQP solver (NZNN-TVQP).
• Investigate the performance of four variations of the proposed NZNN-TVQP solver.
• Evaluate the solver's effectiveness on both simulated and real-world problems.

Main Methods:

• Development of a new adaptive neutrosophic logic/fuzzy zeroing neural network (NZNN) technique.
• Integration of a neutrosophic logic/fuzzy adaptive penalty function.
• Testing and comparison of four NZNN-TVQP solver variations.

Main Results:

• All four variations of the NZNN-TVQP solver demonstrated remarkable performance.
• The solver achieved success in two simulation tests.
• Effective application to two real-world portfolio selection problems was confirmed.

Conclusions:

• The proposed NZNN-TVQP solver is a highly effective tool for addressing time-varying quadratic programming.
• The advancements in neutrosophic logic and fuzzy adaptive techniques enhance TVQP solving capabilities.
• NZNN-TVQP offers a robust solution for complex optimization tasks in various domains.