Hi, I haven't used Clementine so I can't say nothing about that but when choose a software depends mostly of in which you have more experience in use, because in the end the most of them are similar

As you may know, implementation of software where you are aware of the processes and are able of modifying scripts in a problem-dependent way is much more desired for a professional research.

Therefore among these, I always prefer using MATLAB, unless there is a particular problem that a) MatLab lacks a proper tool to solve the problem, and b) I have not enough knowledge/time to develop my own script.

So the answer is case dependent, but generally speaking, I would rather work with my own scripts! 

If you are dealing with very large data-base (big-data such as high resolution image data) "R" software would be perfect .

Otherwise, MATLAB could be a more convenient option since it is much user-friendly (at least to me) while let you adjust/customized the tools based on your specific problem.

Weka is a free open-source software that is mainly suitable for classification problems. Another interesting feature is that Weka can be easily integrated to Java programs (you can write your own program to use the algorithms implemented in Weka).

Clementine is a commercial tool, which is probably easier to use in comparison with Weka.

Dear Aliyeh,

Weka is really efficient to develop robust data-driven JAVA applications: it provides a well range of state-of-the art algorithms for data science. Additionaly, it is used in Pentaho, a well-known data analysis platforms: http://www.pentaho.com/.

Kind regards,

Olivier PARISOT