Comments
There's unfortunately not much to read here yet...
Follow the full discussion on Reddit.
The goal of the project is to let developers write automated browser (ie selenium) tests that have no specific DOM knowledge of the target application. The idea is to reduce the need to rewrite these tests when the underlining DOM implementation changes. To accomplish this an RL agent trained on the Miniwob++ dataset to map user specified actions to a series of DOM events. The agent is exposed through a familiar `ActionChains` interface that ques up actions to be performed in a browser.
There's unfortunately not much to read here yet...
Ever having issues keeping up with everything that's going on in Machine Learning? That's where we help. We're sending out a weekly digest, highlighting the Best of Machine Learning.
Discover the best guides, books, papers and news in Machine Learning, once per week.