Machine Learning?

I am reading

Brink, H., Richards, J. W., & Fetherol, M. (2017). Chapter 5: Basic feature engineering Real-World Machine Learning. Shelter Island, NY: Manning Publications.

In Section 5.3.3  "Real-World feature selection example," I came across this paragraph:

For a great use case of feature engineering and feature selection in the real world, let's look at an example from science.
Your task is to find real supernova events from huge astronomical images among a large number of so-called bogus events (events that look real but aren't).   ***.

Einstein@Home already has millions of labeled examples of gamma-ray pulsars and/or gravitational waves that it is seeking.  Almost certainly, there are many more features of these two phenomenon that astronomers are aware of but haven't thought of testing for, perhaps because of the difficulty.  The existing labeled instances could be augmented with these features using feature engineering (a combination of available features that have not yet been used and domain knowledge).  Then several machine learning algorithms and feature selection procedures could be used in various combinations systematically chosen to determine the model judged to be most predictive of finding an event occurrence on new data.

Many companies are making real breakthroughs in production and marketing processes by applying machine learning to them.  I was wondering if anyone at Einstein@Home has thought of applying machine learning to their goals?