Projects > Discovery Challenge 2004 > Data transformation

Characteristics

A characteristics is a value or a group of values which describes a time series in a specified way.

In the TimeTransf module you must specify which computing function you want to use (for example: average, trend, ) and the variable (an attribute in the original analysed data) for which you want to compute the characteristics. The result value of the characteristics is written to a result matrix for each patient. The result matrix is saved to the original analysed data. Later in data-minig process you use values of these characteristics instead of the original time series.

Example of a characteristics:

Printable version of page PDF version of page

Projects > Discovery Challenge 2004 > Data transformation

Mail to: webmaster