Algorithms from the field of pattern recognition are used throughout industry and academia. In particular machine learning algorithms have become an established technique in modern experimental physics, for instance for the discrimination between desired measurements and background. In recent years, the concept of deep learning has gained considerable attention in the machine learning community and beyond, achieving record breaking performance for a multitude of tasks. This talk introduces some basics of pattern recognition and a few of the most popular approaches in deep learning, namely convolutional neural networks (CNNs), recurrent neural networks and autoencoders. Results of first applications of CNNs for the KM3NeT neutrino telescope are also presented. They have been produced using the Keras framework with TensorFlow, the open source tool from Google.