EA-EMA Optimization Applied to Killer Sudoku Puzzles

AbstractThis paper studies techniques to reduce the search space when an optimizer seeks an optimal value. The paper promotes a new Evolutionary Algorithm (EA) mutation technique called the “Exponential Moving Average” algorithm (EMA). The paper compares its performance to two other similar Computational Intelligence (CI) algorithms to solve a multi-dimensional problem which has a large search space. Testing of the various algorithms is performed against the same Killer Sudoku puzzle and the results compared. The EMA-based solver outperforms an ordinary Evolutionary Algorithm based solver and a “Mean-Variance Optimization” (MVO) solver