Poll | ||||||||||||||||||||||||||||||||||
|
Note from the Editor:
Percentages are given relative to the number of voters. The average number of methods per voter was 3.2.
Here are the results of 2005 KDnuggets Poll on Data mining/analytic techniques.
The percentages in 2005 are not directly comparable since they were computed relative to the number of votes, and not to the number of voters (as in 2006), but we can compare rankings.
We see that the top 4 methods are the same
The methods that significantly gained/lost in ranking were:
Method | 2006 Rank | 2005 rank | Gain/Loss |
---|---|---|---|
Association Rules | 5 | 7 | +2 |
SVM | 7 | 9 | +2 |
Sequence/Time series analysis | 9 | 11 | +2 |
Nearest Neighbor | 11 | 8 | -3 |
Neural Nets | 8 | 6 | -2 |