1 Temporal Rules and Temporal Decision Trees : A C 4 . 5 Approach

@inproceedings{Karimi20011TR,
  title={1 Temporal Rules and Temporal Decision Trees : A C 4 . 5 Approach},
  author={Kamran Karimi},
  year={2001}
}
Decision trees are useful tools for classification and prediction purposes, and have been applied mostly to data that is void of any explicit notion of time. This covers many application areas where the data is about populations of the same entities, but is not suitable for cases where there is a temporal relation among the data. One case is when we are gathering data about a single system over time. In this paper we use a toy domain to show how flattening such data by merging consecutive… CONTINUE READING