In this chapter, we learned how to use TensorFlow Core, TensorFlow Estimators, and Keras packages in R to build and train machine learning models. We provided a walkthrough of the MNIST examples from RStudio and provided links for further documentation of the TensorFlow and Keras R packages. We also learned how to use the visualization tool TensorBoard from within R. We also introduced a new tool from R Studio, tfruns, which allows you to create reports for multiple runs, analyze and compare them, and save them locally or publish them.
The ability to work directly in R is useful because plenty of production data science and machine learning code is written using R, and now you can integrate TensorFlow into the same codebase and run it within the R environment.
In the next chapter, we shall learn some techniques for debugging the code for building and training TensorFlow...
