Extrapolation and learning equations

Georg Martius, Christoph H. Lampert

Nov 02, 2016 (modified: Jan 09, 2017) ICLR 2017 conference submission readers: everyone
  • Abstract: In classical machine learning, regression is treated as a black box process of identifying a suitable function from a hypothesis set without attempting to gain insight into the mechanism connecting inputs and outputs. In the natural sciences, however, finding an interpretable function for a phenomenon is the prime goal as it allows to understand and generalize results. This paper proposes a novel type of function learning network, called equation learner (EQL), that can learn analytical expressions and is able to extrapolate to unseen domains. It is implemented as an end-to-end differentiable feed-forward network and allows for efficient gradient based training. Due to sparsity regularization concise interpretable expressions can be obtained. Often the true underlying source expression is identified.
  • TL;DR: We present the learning of analytical equation from data using a new forward network architecture.
  • Conflicts: ist.ac.at
  • Keywords: Supervised Learning, Deep learning, Structured prediction