A substantial part of Orange is devoted to machine learning methods for classification. These methods start from the data that incorporates classified instances (e.g., a collection of attribute values that are classified to a certain class), and attempt to develop models that would, given the set of attribute-values, predict a class for such instance.
This part of tutorial will show you how to use several of Orange's classification methods. We start with a simple ones (naive Bayes) and then demonstrate the use of several other as well. Another issue we cover is assessing the predictive quality of resulting classifiers. Orange was meant as a versatile tool that would also allow researchers to develop their own methods (or use a combination of existing methods). It is here where we think that Orange in combination with Python scripting really shines. Because Python may prove as a powerful tool to prototype and test your ideas, we include a special lesson on building the classifiers in Python.