[mlpack-svn] [MLPACK] #245: Apply tree traverser abstractions to emst

MLPACK Trac trac at coffeetalk-1.cc.gatech.edu
Wed Aug 15 18:59:56 EDT 2012


#245: Apply tree traverser abstractions to emst
-------------------------+--------------------------------------------------
 Reporter:  rcurtin      |        Owner:                      
     Type:  enhancement  |       Status:  new                 
 Priority:  major        |    Milestone:  mlpack 1.0.3        
Component:  mlpack       |     Keywords:  emst, tree traverser
 Blocking:               |   Blocked By:                      
-------------------------+--------------------------------------------------
 We need to rewrite the emst algorithm to work on arbitrary trees, so we
 must retool RangeSearch to use the new tree traversers.  We will have to
 make a DTBRules class (or something) which implements the following
 functions:

  * `double BaseCase(const size_t queryIndex, const size_t
 referenceIndex);`

  * `double Score(const size_t queryIndex, TreeType& referenceNode);`
  * `double Score(const size_t queryIndex, TreeType& referenceNode, const
 double baseCaseResult)`;
  * `double Rescore(const size_t queryIndex, TreeType& referenceNode, const
 double oldScore)`;

  * `double Score(TreeType& queryNode, TreeType& referenceNode);`
  * `double Score(TreeType& queryNode, TreeType& referenceNode, const
 double baseCaseResult);`
  * `double Rescore(TreeType& queryNode, TreeType& referenceNode, const
 double oldScore);`

  * `void UpdateAfterRecursion(TreeType& queryNode, TreeType&
 referenceNode);`

 See `NeighborSearchRules` for an example implementation.  Bill, I cc'ed
 you not because I need you to solve the problem (though of course I don't
 mind if you do) because it's your method and the eventual solution may be
 interesting to you.

-- 
Ticket URL: <https://trac.research.cc.gatech.edu/fastlab/ticket/245>
MLPACK <www.fast-lab.org>
MLPACK is an intuitive, fast, and scalable C++ machine learning library developed by the FASTLAB at Georgia Tech under Dr. Alex Gray.


More information about the mlpack-svn mailing list