In this thesis the k-means clustering and a neural network is developed and used to classify capacitance data from multi phase flow in a horizontal tube. Theoretical background for the unsupervised machine learning ...