A RANDOMIZED ALGORITHM FOR SOLVING n-DIMENSIONAL SUDOKU PUZZLE WITH CONTINUOUS DIFFICULTY LEVEL
This article consists of two guidelines, the first is the structuring of a method for Sudoku puzzles using random matrices generated by the Bernoulli distribution with parameter p. This allows introducing a continuous measure of the difficulty of a Sudoku generated in this way, unlike the levels of difficulty of the discrete case (easy, light, medium, difficult and hard). The second part consists in designing an algorithm for solving Sudoku puzzles generated by the random matrices method. The proposed algorithm is based on the theory of sets and functions.
Sudoku, Bernoulli distribution, permutations, Latin squares, design of experiments.