DMProject

Just a small course project.

Structure

The data can be downloaded from here.

Folders:

• data: put all the datasets, listed as follows:

  • kddcup.data_10_percent_corrected the main dataset(for now)
  • kddcup.names the names file from the dataset.

• dmproject main code folder

• tools some small code files for something like analysing the dataset.

notice: In .gitignore file, ignore the data/ folder, cause it's really large.

Roadmap (a sketch)

Algorithms:

• KNN method.
• Hierarchical clustering method.
• Other density based method? (maybe)

Pre-processing methods:

• normalization methods?
• feature selection? (maybe)