Algorithms for Direct 0-1 Loss Optimization in Binary Classification

By Tan Minh Nguyen, Scott Sanner. 2013

125 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.

In Statistical Learning Theory

Citations in the corpus

Learning Explainable Decision Rules via Maximum Satisfiability (2020, background): “Owing to the 0-1 loss function in (6), ﬁnding direct solutions of (3)-(5) is known to be NP-hard [20].”
The landscape of empirical risk for nonconvex losses (2016, background): “First of all, several empirical studies [4, 21, 15] demonstrate an improved classification error using non-convex losses.”
Fairness in Machine Learning: A Survey (2020, unknown): “Note that sufficient data is required for each subgroup for this method of sampling to not negatively affect performance and fairness, as shown by [9], and that outliers can be problematic [45, 211].”
Learning to Teach with Dynamic Loss Functions (2018, methodology): “…boosting the performances of real-world tasks, either by approximating the non-smooth task-speciﬁc objective such as 0-1 accuracy in classiﬁcation [41], NDCG in ranking [52], BLEU in machine translation [48, 3] and MAP in object detection [22, 49], or easing the optimization process of the…”
Fairness without Harm: Decoupled Classifiers with Preference Guarantees (2019, background): “Such violations can stem from a lack of data for some groups, or the fact that surrogate losses may not be robust to outliers (e.g., points belonging to a heterogeneous subpopulation; Brooks, 2011; Nguyen & Sanner, 2013).”
Generative Adversarial Privacy: A Data-Driven Approach to Information-Theoretic Privacy (2018, methodology): “This is achieved by considering a 0-1 loss function (Nguyen & Sanner, 2013) given by `(h(g(X)), Y ) = { 0 if h(g(X)) = Y 1 otherwise .”
Learning Generative Adversarial RePresentations (GAP) under Fairness and Censoring Constraints (2019, methodology): “This is achieved by considering a 0-1 loss function Nguyen and Sanner [2013] given by”
A Group-Theoretic Framework for Data Augmentation (2019, unknown): “, logistic loss and hinge loss (Wu and Liu, 2007; Nguyen and Sanner, 2013).”
Resilient Linear Classification: An Approach to Deal with Attacks on Training Data (2017, unknown): “, [37]) and it would be promising to design and analyze multiple algorithms in tandem (one to monitor the data, one to learn a classifier).”
Context-Aware Generative Adversarial Privacy (2017, methodology): “This is achieved by considering a 0-1 loss function [63] given by”
Censored and Fair Universal Representations using Generative Adversarial Models (2019, methodology): “This is achieved by considering a 0-1 loss function [Nguyen and Sanner, 2013] given by”