Lessons Learned from Benchmarking Fast Machine Learning Algorithms

Post by cropsieboss » Wed, Jul 26, 2017, 1:22 PM UTC

insulator wrote: post #20

cropsieboss wrote: post #6

It would be interesting if they compared training speed with CatBoost [0]. I remember seeing a paper where they managed to avoid getting stuck in local optimum in terms of number of learners, and the more trees you add better the result. Logloss results seem to confirm there's a superior tree algorithm going on there in CatBoost. [0]: https://catboost.yandex/

here is a benchmark for catboost: https://github.com/szilard/GBM-perf/issues/4