The purpose of my thesis was to propose four novel implementations of univariate kernel density estimation based on TensorFlow, TensorFlow Probability and zfit, which incorporate insights from recent papers to decrease the computational complexity and therefore runtime. The newly proposed implementations were then compared to the state of the art of kernel density estimation in Python and shown to be competitive. By leveraging TensorFlow’s graph based computation, the newly proposed methods to calculate a kernel density estimate can benefit from parallelization and efficient computation on CPU/GPU.