During the past two decades, the number of volumetric seismic attributes have increased to the point in which interpreters are overwhelmed and cannot analyze all the information available. Principal Component Analysis (PCA) is one of the best-known multivariate analysis technique, and decomposes the input data into lower statistics mathematically uncorrelated components. Unfortunately, while these components mathematically represent the information in the multiple input data volumes using a smaller number of volumes, they often mix rather than separate geologic features of interest. To address this issue, I implement and evaluate a relatively new unsupervised multi-attribute technique called Independent Component Analysis (ICA), which based on higher order statistics, separates multivariate data into independent subcomponents. I evaluate my algorithm to study the internal architecture of turbiditic channel complexes present in the Moki A sands Formation, Taranaki Basin, New Zealand. I input twelve spectral magnitude components ranging from 25 to 80 Hz into the ICA algorithm and plot three of the resulting independent components against an RGB color scheme to generate a single volume in which different colors correspond to different seismic facies. The results obtained using ICA proved to be superior to the obtained using PCA. Specifically, using ICA I obtain independent components that have better resolution and better separation between geologic features and noise compared to uncorrelated components obtained using PCA. Moreover, with ICA, I am able to geologically analyze the different seismic facies and relate them to sand-prone and mud-prone seismic facies associated with axial and off-axis deposition and cut-and-fill architectures.