In early March, the Bay Area useR Group was able to hold an R and TensorFlow mini-conference on Google’s new Sunnyvale campus. Pete Mohanty, a Stanford researcher and frequent BARUG speaker, lead off with a talk on his recent kerasformula package, which allows R users to call a keras-based neural net with R formula objects. Pete’s slides show an example of using using a regression-style formula with the kerasformula::kms() function to fit a sequential TensorFlow model.

J.J. Allaire, RStudio’s founder and CEO, spoke for over an hour, delivering a polished and comprehensive presentation that ranged from big-picture vistas illuminating the merits and limitations of the Deep Learning methodology to the deep details of R based TensorFlow models. We were not able to record J.J.’s BARUG talk, but it was similar to his presentation at the January RStudio conference in San Diego, which is well worth watching.

We do have the slides for J.J.’s BARUG presentation. They comprise a comprehensive overview of Deep Learning and the TensorFlow Technology behind it. While most slide presentations are little more than collections of mnemonics, lifeless without the animation of the speaker, J.J.’s 113-slide deck stands on its own. Replete with links to current research papers and references that guide a reader to the frontiers of Deep Learning applications, is a very credible introduction and study guide. Because 113 slide are quite a bit to get through in one sitting, especially if you start following interesting links, I offer the following gloss: