[mlpack-git] [mlpack/mlpack] Multiprobe LSH Implementation (#657)

[Yannis Mentekidis]

First version of Multiprobe LSH. It is possible GetAdditionalProbingBins() can be further optimized.

I ended up using the initial code that allocated more memory and used find() to create the candidate set - I could change that to allocating less memory and using unique() instead, as I proposed in PR #623.

More eyes on this might be beneficial so I am submitting a PR.


You can view, comment on, or merge this pull request online at:

  https://github.com/mlpack/mlpack/pull/657

-- Commit Summary --

  * Implements GetAdditionalQueryBins
  * First working version, requires more testing
  * First working version, requires more testing
  * Fix minor bug causing freeze when too many probes were requested
  * Adds ComputeRecall and changes lsh_main.cpp to call it if truth file is specified
  * Fixes regression in ReturnIndicesFromTables that caused segfaults
  * Fixes regression in ReturnIndicesFromTables that caused segfaults
  * Prevents perturbation sets with invalid dimensions to be created
  * Adds multiprobe test
  * Adds multiprobe test
  * Documents ComputeRecall
  * Style Fixes
  * Style Fixes
  * Merge branch 'master' of https://github.com/mlpack/mlpack into MultiprobeLSH

-- File Changes --

    M src/mlpack/methods/lsh/lsh_main.cpp (22)
    M src/mlpack/methods/lsh/lsh_search.hpp (31)
    M src/mlpack/methods/lsh/lsh_search_impl.hpp (352)
    M src/mlpack/tests/lsh_test.cpp (67)

-- Patch Links --

https://github.com/mlpack/mlpack/pull/657.patch
https://github.com/mlpack/mlpack/pull/657.diff

---
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/657
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160529/ae1bf2d9/attachment.html>