In recent years the interaction between beta-cyclodextrin and cholesterol has been an area of much research. This is mainly due to the assumption that beta-cyclodextrin has a positive effect on cholesterol based diseases. In this project different unsupervised learning tools are used to study the interaction between the two molecules, beta-cyclodextrin and cholesterol. The examined data set consist of computer simulations of the above stated interaction, which provides 1001 different images. It is desired to apply different clustering methods to achieve a representation of the data set with a reduced number of images. This is shown to be feasible.