DATA MINING
Desktop Survival Guide by Graham Williams |
|||||
Number of Algorithms |
Users and publicists will often quote the number of algorithms available within a data mining package as a measure of how good the package is. This is not really a good measure, since it is more important to have the right algorithms, and a small number so as not to confuse the new data miner. Yet, it is also very useful to have access, as required, to the latest collection of algorithms so the data miner can explore them as they gain experience.
The Java-based Weka package has an extensive number of learners implemented, and perhaps the largest number of any data mining tool (apart from R, which in fact also ``includes'' Weka).
R has an extensive range of packages, many of which provide support for the data miner. But, on top of that, The RWeka package provides access from within R to all of the algorithms in Weka!