As mentioned in Chap. 1, appropriate rule representation is necessary in order to improve model efficiency and interpretability. This chapter introduces three techniques for representation of classification rules namely, decision trees, linear lists and rule based networks. In particular, these representations are illustrated using examples in terms of searching for firing rules. These techniques are also discussed comparatively in terms of computational complexity and interpretability.
|Title of host publication||Rule Based Systems for Big Data|
|Subtitle of host publication||A Machine Learning Approach|
|Number of pages||12|
|ISBN (Print)||9783319236957, 9783319370279|
|Publication status||Published - 17 Sept 2015|
|Name||Studies in Big Data|