Description of Invention
Simulated Kalman Filter (SKF) and its mutation-based variant, SKF-MUT, are optimization algorithms inspired by the Kalman filter. SKF utilizes Kalman equations for solution updates, with random elements promoting exploration and exploitation. SKF-MUT addresses premature convergence by introducing mutation: a mutated solution (Xmut) is generated post-estimation, potentially replacing the best solution (Xbest) if it yields better fitness. Otherwise, it replaces a random agent. SKF-MUT enhances optimization performance, exemplified by a 0.6% improvement in classifying 11 tumors in DNA microarray data. SKF-MUT offers promising solutions for complex optimization problems, leveraging the power of Kalman-based estimation and controlled mutation to achieve optimal results.