Accurately describing the gamut of a color device is the basis for gamut mapping, device color characterization and device gamut volume prediction. There are many ways to describe gamut boundary in the past and the methods can be used in combination with each other for the more accurate and effective gamut boundary description. However, it is difficult to find a commonly used method after introducing the way of color space segmentation. In this paper, we reorganize the existing gamut boundary description techniques according to purpose and method. We also propose a new simple approach of predicting gamut boundary. This new approach uses a machine learning based Radial Basis Function Network(RBFN) that can simplify the gamut boundary description process. This simple method can directly predict the desired gamut boundary description.