On Issues of Instance Selection

By Huan Liu, Hiroshi Motoda. 2002

224 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Evaluation & Benchmarks

How researchers see this paper

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

In Artificial Intelligence

In Evaluation & Benchmarks

Citations in the corpus

Understanding Dataset Difficulty with V-Usable Information (2021, neutral): “Estimating instance difficulty is also evocative of instance selection for active learning (Lewis & Catlett, 1994; Fu et al., 2013; Liu & Motoda, 2002); however these estimates could change as the dataset picks up new instances.”
WoodScape: A Multi-Task, Multi-Camera Fisheye Dataset for Autonomous Driving (2019, methodology): “One frequently used approach is instance selection [40, 35, 26].”
FSD50K: An Open Dataset of Human-Labeled Sound Events (2020, background): “Under the assumption that not all training examples are equally informative (which is the basis for disciplines like instance selection [125]), this may occur because the content transferred to val includes some highly informative or hard examples.”
Information-Theoretic Measures of Dataset Difficulty (2021, background): “Estimating instance difﬁculty is evocative of instance selection for active learning (Fu et al., 2013; Liu and Motoda, 2002).”
A review of instance selection methods (2010, background): “Some authors (Bezdek and Kuncheva 2001; Liu and Motoda 2002; Spillmann et al. 2006; Caises et al. 2009) have stated the idea of using clustering for instance selection; this idea consists of: after splitting T in n clusters, the selected instances will be the centers of the clusters.”
Evolutionary stratified training set selection for extracting classification rules with trade off precision-interpretability (2007, methodology): “This process is carried out by means of instance selection algorithms [6] [7] where, instead of removing the irrelevant instances, the most representative ones are selected.”
A Survey on Evolutionary Instance Selection and Generation (2010, background): “• Selecting instances (Liu & Motoda, 2001; Liu & Motoda, 2002), reducing the number of rows in a data set.”
Addressing Missing Labels in Large-Scale Sound Event Recognition Using a Teacher-Student Framework With Loss Masking (2020, background): “…that a tiny portion of labels is troublesome and it is moderately affecting classiﬁer performance, a concept which is basis for disciplines like instance selection, where it is assumed that not all training examples are equally informative, some of them being redundant and some being harmful [27].”
Big data preprocessing: methods and prospects (2016, background): “Nowadays, instance selection is perceived as necessary [32].”