Industry benchmarks to assess data mining tools

Question

Hi

I'm looking for data mining tools for a project and in line with that I have put up another post in SO. I'm currently looking at different tools and am wondering whether any industry benchmark exists to asses different data mining tools so that I can refer it do a better evaluation of tools.

Please let me know if any such benchmark exists or it just boils down to picking the tool which suits your needs ...

cheers

Answer 1

Here are some standard sets of test data to benchmark accuracy performance:

• UC Irvine - Knowledge Discovery in Databases (KDD) Dataset Archive

• UToronto - Data for Evaluating Learning in Valid Experiments (DELVE)

Answer 2

For evaluations of data mining tools on features/etc, how about these links:

• A Comparison of Leading Data Mining Tools (1998) - dated but pretty informative
• Magic Quadrant for Customer Data-Mining Applications (2008) - a bit non-technical

Answer 3

This site: http://clopinet.com/challenges/ contains tons of datasets for machine learning techniques. The best part is that you can also compare your performance to that of others.