I need to test various famous classification methods like kNN, ID3 and ... on a huge data-set of a project, and choose one for future use.
I have no limitation on language but performance and readable code both in learning and classification phase are ver开发者_运维百科y important.
therefore, I'm looking for a good library with following features:
- includes various classification methods
- high performance
- easily usable
any suggestions?
You shuold take a look at PyBrain, a great machine learning module for Python. Can't tell you much about it, because I never really used it (just read about it and looked at several projects solved with it), but it seems to be very good.
You may also want to take a look at this list of Python machine learning modules: http://web.media.mit.edu/~stefie10/technical/pythonml.html
Or at this SO question: Is there a recommended package for machine learning in Python?
Try scikit learn - it is written in python and has variety classifiers and also metods for testing them.
Take a look at RapidMiner which comes with a Java-API and graphical tools for data mining. The community edition is free, I think.
I used the predecessor of this tool/library as a student but do not have professional experience with it, though.
精彩评论