Changes between Version 15 and Version 16 of GSoC/Ideas


Ignore:
Timestamp:
03/10/11 16:00:44 (4 years ago)
Author:
lanz
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • GSoC/Ideas

    v15 v16  
    2828 
    2929Useful skills: Python. The candidate should be familiar with statistics and computation with matrices (numpy). 
     30 
     31Level from 1 (beginner) to 5 (professional): 3 
     32 
     33== More state-of-the-art regression algorithms == 
     34 
     35Find and research state-of-the art regression algorithms found in other packages (R, Weka, ...) and choose the most important representatives not found in Orange. Then reimplement them in Orange. Another option is to add just the wrappers for external libraries if they are open-source and compatible enough. 
     36 
     37Useful skills: Python. Probably also statistics and computation with matrices (numpy). 
     38 
     39Level from 1 (beginner) to 5 (professional): 4 
     40 
     41== Improve k-Nearest Neighbors == 
     42 
     43For n training and m test examples standard kNN checks the distances to all n training for each test examples. The time complexity is thus O(n*m). 
     44With a smarter implementation this can be improved. 
     45 
     46Useful skills: Python and C. 
    3047 
    3148Level from 1 (beginner) to 5 (professional): 3