Changes between Version 31 and Version 32 of Parallel-DT


Ignore:
Timestamp:
Jan 18, 2010, 9:17:10 PM (14 years ago)
Author:
andrei.minca
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • Parallel-DT

    v31 v32  
    2727 * '''Syncronous Tree Construction  - Depth First Expansion Strategy''' - the one that we implemented
    2828
    29 In this approach, all processors construct a decision tree syncronously by sending and receiving class distribution information of local data.
     29In this approach, all processors construct a decision tree syncronously by sending and receiving class distribution information of local data. Major steps for the approach:
     30   - select a node to expand according to a decision tree expansion strategy (eg Depth-First or Breadth-First), and call that node as the current node. At the beginning, root node is selected as the current node
     31   - for each data attribute, collect class distribution information of the local data at the current node
     32   - exchange the local class distribuition information using global reduction among processors
     33   - simultaneously compute the entropy gains of each attribute at each processor and select the best attribute for child node expansion
     34   - depending on the branching factor of the tree desired, create child nodes for the same number of partitions of attributes values, and split training cases accordingly
    3035
    3136[[Image(SyncronusTreeConstruction-DepthFirstExpansionStrategy.jpg)]]