application problem. The question arises – which one should be
used?
Appropriateness of a suitable technique to a data set should be
based on certain factors such as characteristics of the data set,
strength and weakness of each individual method under
consideration, availability of resources, usefulness of the
concluded result and some previous but recent comparative
studies on effectiveness of various methods.
Recently according to several researchers, an ensemble of data
analysis methods (classifiers) can often yield more accurate
results than any individual method (classifier) alone. The
integration of classifiers can be made statically, as dividing a
problem into sub-problems, or dynamically, taking into account
the variables of the new pattern during analysis.
Do'stlaringiz bilan baham: