An instance level analysis of data complexity

By Michael R. Smith, Tony R. Martinez, Christophe G. Giraud-Carrier. Machine-mediated learning, 2014

437 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/9cbfbf6508f591555a898f6ebca4d4608fd02bea

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

Citations in the corpus

Meta-QSAR: a large-scale application of meta-learning to drug design and discovery (2017, methodology): “Recent research on ﬁnding interesting ways to measure data characteristics includes instance-level complexity (Smith et al., 2014a), measures for unsupervised learning (Lee and Giraud-Carrier, 2011), and discretized meta-features (Lee and Giraud-Carrier, 2008).”
Striking the Right Balance With Uncertainty (2019, neutral): “1.1 75.210.6 88.26.0 76.911.4 Under-samp. Clustering [59] 75.7 78.312.2 75.710.3 75.96.9 87.06.8 74.812.2 Neighborhood Cleaning [26] 76.7 79.210.4 76.718.6 75.98.6 85.710.2 73.218.1 Instance Hardness [45] 67.8 76.3 ble 1. However, here our main goal is to analyze if the un-19.8 67.818.7 67.610.8 79.”
A Cluster-Based Boosting Algorithm for Bankruptcy Prediction in a Highly Imbalanced Dataset (2018, methodology): “Instance Hardness (IH) was proposed by Smith, Martinez, and Giraud-Carrier [30] in 2014 and is only The smaller the value of p ( y i | x i , h ), the less correct h is. used for binary classification problems.”
Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning (2016, methodology): “The implemented methods perform the under-sampling based on diﬀerent criteria such as: (i) random selection, (ii) clustering, (iii) nearest neighbours rule (i.e., NearMiss (Mani and Zhang, 2003)), and (iv) classiﬁcation accuracy (i.e., instance hardness threshold (Smith et al., 2014)).”
How Complex Is Your Classification Problem? (2018, methodology): “Other uses of the complexity measures in the noise identiﬁcation context are: [Smith et al., 2014a, Saéz et al., 2013, Gar´cia et al., 2013, Garcia et al., 2015, 2016].”