On the Maximal Size of Large-Average Submatrices in a Gaussian Random Matrix: Theoretical and Numerical Approaches



Based on the article “On the Maximal Size of Large-Average and ANOVA-Fit Submatirces in a Gaussian Random Matrix” by Sun and Nobel, we investigate the maximal size of submatrices with average of the values of such submatrices more than a fixed positive number in the Gaussian random matrix. We identify the limit behavior of the threshold of the size of submatrices theoretically and numerically. Our principal result is an inconsistency between the results of two approaches and we propose our own analysis from the theoretical and numerical perspectives.

Final Report