Hello there, Today I have a long-awaited announcement---we have released mlpack 3.0.0. This is a big change architecturally for the library and it includes a lot of added functionality as well as other improvements. This has been a huge community effort, and at this point we have over 110 contributors. I think this is really great and I think we have shown as a community that we can work together to produce high-quality machine learning software!
You can get the new release at http://www.mlpack.org/files/mlpack-3.0.0.tar.gz and over the coming days, it will percolate into different package repositories, so you will be able to get it with your favorite package manager (now including pip and conda, since the Python bindings are now released!). As always, here's a shortlist of changes: * Speed and memory improvements for DBSCAN. --single_mode can now be used for situations where previously RAM usage was too high. * Bump minimum required version of Armadillo to 6.500.0. * Add automatically generated Python bindings. These have the same interface as the command-line programs. * Add deep learning infrastructure in src/mlpack/methods/ann/. * Add reinforcement learning infrastructure in src/mlpack/methods/reinforcement_learning/. * Add optimizers: AdaGrad, CMAES, CNE, FrankeWolfe, GradientDescent, GridSearch, IQN, Katyusha, LineSearch, ParallelSGD, SARAH, SCD, SGDR, SMORMS3, SPALeRA, SVRG. * Add hyperparameter tuning infrastructure and cross-validation infrastructure in src/mlpack/core/cv/ and src/mlpack/core/hpt/. * Fix bug in mean shift. * Add random forests (see src/mlpack/methods/random_forest). * Numerous other bugfixes and testing improvements. * Add randomized Krylov SVD and Block Krylov SVD. But this is really not a great overview of everything that has changed since there has been so much. However I don't want to make this email ten thousand lines long. In any case, if you're interested in all the changes, you could look at the full list of commits (which is very long, over three thousand commits!): https://github.com/mlpack/mlpack/compare/mlpack-2.2.5...mlpack-3.0.0 Or the list of merged pull requests: https://github.com/mlpack/mlpack/pulls?utf8=%E2%9C%93&q=is%3Apr+is%3Amerged Thank you so much to everyone who has contributed to the library. I am sure we can all see that without the growing community we have here, none of this is possible. I've written a blog post announcing the release also, which I hope to share in the next few days (please feel free to share it also): http://www.mlpack.org/blog/mlpack-3-released.html In the future, I am sure that there is much that can still be added and improved! We currently have 57 pull requests open and I'll be trying to do my part to make sure some of this great functionality gets into the library. I think also we can expect bindings to be made available for other languages in future releases, and it's likely that additional optimizers will continue to be added too. If you have any issues with the new version, please feel free to file a bug on Github (https://github.com/mlpack/mlpack) or chat in the IRC channel #mlpack on freenode. I hope everyone has a great weekend! I will be relaxing now that this release is done. (For a little while, at least...) -- Ryan Curtin | "He takes such wonderful pictures with r...@ratml.org | his paws." - Radio man _______________________________________________ mlpack mailing list mlpack@lists.mlpack.org http://knife.lugatgt.org/cgi-bin/mailman/listinfo/mlpack
