I’d like to acquaint you with a tremendous resource for keeping current with the latest research in the field of machine learning. Informally known as the pre-print server, arXix.org is the global repository for the fields of Physics, Mathematics, Computer Science, Quantitative Biology, Quantitative Finance and Statistics. A recent tally showed 898,961 papers posted for free public consumption. Recently submitted papers for machine learning can be found HERE.
arXix.org is sponsored by the Cornell University Library and serves as an integral resource for scientist around the world. Many researchers submit papers reporting new results to the pre-print server long before you see them in refereed journals. This means you can see the latest and greatest as soon as the results become available.
I check the machine learning category about once-per-week, but in my previous life as an independent researcher in astrophysics, the pace of papers was much more fierce and I had to check daily (midnight each day was my time to find out what was new with gravitational wave astronomy!).
If you’re serious about this field, I recommend you become familiar with this resource. The papers are written for graduate level understanding and at first not much may make sense, but with a little fortitude and time, important concepts will start to sink in and you’ll feel accomplished in knowing you’re at the forefront of this important field.