In class over the last two weeks we’ve discussed basics of machine learning and non-parametric tests.
We also started our class projects in earnest. My part so far has been to figure out data ingestion and how to convert from simulation units to physical units. This took longer than expected but I think I’ve got it figured out now.
Big Data and Machine Learning Workshop
I attended XSEDE’s Big Data and Machine Learning workshop given by the staff a the Pittsburgh Supercomputing Center (PSC). It was a two day truly excellet workshop that provided and intro to Spark and various machine learning algorithms such as K-nearest neighbors, Alternating Least Squares, neural nets (including deep learning, convolutional neural nets, etc.). It was a really great workshop, as are all the PSC/XSEDE workshops and I highly recommend it. The recorded lectures can be found on their youtube channel, though at the time of writing the most current Big Data and ML workshop is not yet posted.
I implemented the single wave initial conditions I discussed in Classes and Bugfixing 6. Currently the contact discontinuity appears to work except for some issues at the boundaries. I’m not sure what that is yet as the periodic boundary conditions appear to be ok, I need to check with a pure hydro wave to see if that works.
- Finished my Statement of Purpose for Rensselaer summer school application
- Finished the MHD review article Spruit 2017
- Minor CV updates