High-dimensional data are common in many domains, and dimensionality reduction is the necessary to cope with the curse-of-dimensionality. This phenomenon states that an enormous number of samples is required to perform accurate predictions on problems with high dimensionality. Dimensionality reduction, which extracts a small number of features by removing irrelevant, redundant, and noisy information, can be an effective solution. Different statistical methods for dimensionality reduction have been proposed in recent years and various research groups have reported contradictory results when comparing them. The commonly used dimensionality reduction techniques include supervised approaches such as Linear Discriminant Analysis (LDA), and unsupervised ones such as Principal Component Analysis (PCA), and additional spectral and manifold learning methods. When class labels are available, the supervised approaches such as LDA are generally more effective than the unsupervised ones like PCA for classification. This paper aims at the review of two most widely used dimensionality reduction techniques, PCA and LDA. Based on this a way ahead will be presented to facilitate research and development in sediment classification.