I have to select a DM research paper which presents an interesting novel algorithm (it can be
extended from an existing method).
2) Implement the algorithm as a DM tool for the selected application. You can
choose a programming language, but it must be compliable on either hector or
3) Find a real world application (database) and develop an application scenario
based on the data and the algorithm in 1). There are many data sources on the
Web, such as the example provided in Data/README.
4) Identify what data preparation is needed, and present the work you have done in
your report. (For a classification task, the selected data set needs to be divided
into two subsets for training and testing respectively.)