Github user avulanov commented on the pull request:
https://github.com/apache/spark/pull/1290#issuecomment-62086344
@witgo I have implemented ANN in matrix form for Spark with breeze a while
ago: https://github.com/avulanov/spark/tree/neuralnetwork. We tested it with
@bgreeven and it was less effective than loop-based due to overhead of matrices
rolling/unrolling. The latter is needed because `Gradient` class does not allow
passing matrices. I also tried plugging native BLAS/LAPACK into breeze but it
didn't deliver better performance probably because loops are well-optimized in
case of vector-matrix multiplication. Do you think we still should use
vectorized form for better code readability?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
