Open Peer Review. Open Publishing. Open Access. Open Discussion. Open Directory. Open Recommendations. Open API. Open Source.
Comparison of two gradient computation methods in Python
Sri Hari Krishna Narayanan, Paul Hovland, Kshitij Kulshreshtha, Devashri Nagarkar, Kaitlyn MacIntyre, Riley Wagner, Deqing Fu
Oct 28, 2017 (modified: Oct 28, 2017)NIPS 2017 Workshop Autodiff Submissionreaders: everyone
Abstract:Gradient based optimization and machine learning applications require the computation of derivatives. For example, artificial neural networks (ANNs), a widely used learning system, use backpropagation to calculate the error contribution of each neuron after a batch of data is processed. Languages such as Python and R are are popular for machine learning. Therefore, there has been interest in the last few years to build tools for Python and R to compute derivatives.
TL;DR:Compares the performance and features of two automatic differentiation tools, autograd and ADOL-C