Projects > Discovery Challenge 2004 > Data transformation > Time series

 

Time series

Control examinations of a patient forms time series. For example there is time series on attribute weight (HMOT), which shows us how the patient's weight goes in time. To prepare the data for use in data-mining process we need to transform this time series into so called characteristics, which tells us for example: the patients average weight for the period of time, how many kilograms is the patient heavier or lighter every year or in which year the patient began to batten.

Time series can be divided into the subgroups according to the Table 1.

Table 1: Groups of attributes in the control examination
Group of time series Number of used attributes
Identification data 0
changes 8
sickness leave 1
A2 questionnaire 16
physical examination 10
biochemical examination 18

Printable version of page    PDF version of page

Projects > Discovery Challenge 2004 > Data transformation > Time series

 

Mail to: webmaster