A critical review of multi-objective optimization in data mining: a position paper

225 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Supersparse linear integer models for optimized medical scoring systems (2015, background): “on, Bartlett et al. 2003) and sparsity (i.e. by minimizing a convex proxy measure, such as the ‘ 1-norm). Such approximations may result in a poor trade-o between accuracy and sparsity (Bratko, 1997, Freitas, 2004). Linear classiers that minimize surrogate loss functions, for instance, are not robus”
Comprehensible classification models: a position paper (2014, methodology): “We favor instead a more principled approach for coping with the accuracycomprehensibility trade-off, such as multi-objective optimization based on Pareto dominance [25], [38] or lexicographic optimization [25], [42].”
Knowledge actionability: satisfying technical and business interestingness (2007, methodology): “…of pattern mining in traditional data mining techniques such as association rules (Liu et al., 2000), multi-objective optimisation in data mining (Freitas, 2004), role model-based actionable pattern mining (Wang et al., 2006), cost-sensitive learning (Domingos, 1999) and postprocessing (Yang et…”
A Survey of Evolutionary Algorithms for Decision-Tree Induction (2012, background): “For a critical review of the pros and cons of each multiobjective strategy that is discussed in this section, see [30].”