.png){kind=icon}
.png){kind=icon}
Return
Simplification Oriented Data and Pattern Transformations Vs. Attribute Importance and Rule-Based Classifier Performance
.png){kind=icon}
PRE
AI
DOI：10.1007/978-3-032-13869-9_8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Simplification is a popular trend in most activities and areas, including computer science and machine learning. It supports a deeper understanding of application domains and helps with generalisation while reducing computational and storage costs. The paper presents research in which a simplifying procedure in the form of discretisation was applied to the input data and patterns discovered by the Dominance-Based Rough Set Approach. The influence of transformations was studied from the perspective of attribute importance estimated by rankings and performance of rule-based classifiers simplified by ranking-driven rule filtering. The experimental results show many cases of improved predictions and enhanced interpretability due to the employed processing.
Keywords:
Relevance
Ranking
Discretisation
Rule filtering
DRSA
Journal
E
IF:
0
Papers:
27
Citations:
0