ABSTRACT Name : Joshua Casey Darian Gunawan Program : Ilmu Komputer Title : Submatrix Selection for SVLT under Sparse Singular Vector Assumption Matrix rank estimation is a classical problem with many applications and approaches. One of the recent methods of matrix rank estimation is based on singular value thresholding, one of which is called Singular Value Logistic Thresholding (SVLT). In this final project, a method to improve SVLT performance is proposed under the assumption of the observation matrix having sparse singular vectors using submatrix selection. This method is shown to greatly improve the runtime of SVLT while also providing better accuracy for a high-enough Signal-to-Noise Ratio. Keywords: SVLT, matrix rank estimation, sparse singular vectors