Hey all It's almost nostalgic to send out this email as a closing note to my GSoC journey. I can still remember the email I sent out for my draft proposal
What happened during this summer? * Fixed and merged a modified NSGA - II. * Expanded CheckArbitraryFunctionTypeAPI * Extended support of callbacks for Multi-objective use cases. * MOEA/D-DE algorithm. * Portfolio notebook and Rocket Injector notebook * Released a goofy new ensmallen version. * And so much more...! You can find the details here: https://github.com/jonpsy/GSoC-2021-mlpack. I'm proud to say we're on par with some of the hottest Multi-objective libraries and in some cases beat them by a huge margin. After mid evaluation, I developed a liking for Reinforcement learning and dumped my original idea for a "Multi-objective reinforcement learning" algorithm. So I had to learn RL from scratch, and then our RL codebase from scratch and then a MORL algorithm (from scratch ;-D). This ate up a lot of my time, which I don't regret. I hope to continue working on it after GSoC! . A huge thanks to Marcus, Sayan and James for all the support. I'll miss chatting like we used to have on history, about your respective countries and languages. Also a huge shout-out to Ryan, Shrit and all my GSoC peers, you guys played an important role too. Best S
_______________________________________________ mlpack mailing list mlpack@lists.mlpack.org http://knife.lugatgt.org/cgi-bin/mailman/listinfo/mlpack