Theory of Optimizing Pseudolinear Performance Measures: Application to F-measure
원문 초록
Non-linear performance measures are widely used for the evaluation of learning algorithms. For example, $F$-measure is a commonly used performance measure for classification problems in machine learning and information retrieval community. We study the theoretical properties of a subset of non-linear performance measures called pseudo-linear performance measures which includes $F$-measure, \emph{Jaccard Index}, among many others. We establish that many notions of $F$-measures and \emph{Jaccard Index} are pseudo-linear functions of the per-class false negatives and false positives for binary, multiclass and multilabel classification. Based on this observation, we present a general reduction of such performance measure optimization problem to cost-sensitive classification problem with unknown costs. We then propose an algorithm with provable guarantees to obtain an approximately optimal classifier for the $F$-measure by solving a series of cost-sensitive classification problems. The strength of our analysis is to be valid on any dataset and any class of classifiers, extending the existing theoretical results on pseudo-linear measures, which are asymptotic in nature. We also establish the multi-objective nature of the $F$-score maximization problem by linking the algorithm with the weighted-sum approach used in multi-objective optimization. We present numerical experiments to illustrate the relative importance of cost asymmetry and thresholding when learning linear classifiers on various $F$-measure optimization tasks.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Label Consistent Fisher Vectors for Supervised Feature Aggregation
JointRank: Rank Large Set with Single Pass
From Web Search towards Agentic Deep Research: Incentivizing Search with Reasoning Agents
From Chaos to Automation: Enabling the Use of Unstructured Data for Robotic Process Automation
Temporal Information Retrieval via Time-Specifier Model Merging
An analysis of vision-language models for fabric retrieval
토론
이 논문과 연결된 포럼 글이 아직 없습니다.