Dec 7, 2011
0 notes
[We had a breakthrough and came up with the] first learning algorithm I’ve seen that is provably faster than any future single machine learning algorithm. The proof of this is simple: We can output a optimal-up-to-precision linear predictor faster than the data can be streamed through the network interface of any single machine involved in the computation.
Hunch Machine Learning Blog: Hadoop AllReduce and Terascale Learning
About
andy hickl is a data scientist, computational linguist, and entrepreneur who is now head honcho at A.R.O., Inc..