F-测度极大化的Bayes最优性
On the Bayes-optimality of F-measure maximizers
摘要:
F测度最初被引入信息检索中,现在通常用作二进制分类,多标签分类和结构化输出预测等问题的性能度量。优化这一措施是一个统计和计算上的挑战性问题,因为不存在封闭形式的解决方案。采用决策理论的角度来看,本文提供了一个正式的和实验分析的不同方法,最大限度地提高F-措施。我们开始与相关的损失函数,如汉明损失和子集零一损失的贝叶斯风险分析,表明优化这种损失作为替代的F-措施导致高最坏情况下的遗憾。随后,我们进行了类似类型的F-测量最大化算法的分析,表明这种算法是近似的,同时依赖于额外的假设有关的二进制响应变量的统计分布。此外,我们提出了一个新的算法,这不仅是计算效率,而且贝叶斯最优,无论底层的分布。为此,该算法只需要二次(相对于二进制响应的数量)的参数的联合分布。我们通过多标签分类问题的实验说明了所有分析方法的实际性能。
英文摘要:
The F-measure, which has originally been introduced in information retrieval, is nowadays routinely used as a performance metric for problems such as binary classification, multi-label classification, and structured output prediction. Optimizing this measure is a statistically and computationally challenging problem, since no closed-form solution exists. Adopting a decision-theoretic perspective, this article provides a formal and experimental analysis of different approaches for maximizing the F-measure. We start with a Bayes-risk analysis of related loss functions, such as Hamming loss and subset zero-one loss, showing that optimizing such losses as a surrogate of the F-measure leads to a high worst-case regret. Subsequently, we perform a similar type of analysis for F-measure maximizing algorithms, showing that such algorithms are approximate, while relying on additional assumptions regarding the statistical distribution of the binary response variables. Furthermore, we present a new algorithm which is not only computationally efficient but also Bayes-optimal, regardless of the underlying distribution. To this end, the algorithm requires only a quadratic (with respect to the number of binary responses) number of parameters of the joint distribution. We illustrate the practical performance of all analyzed methods by means of experiments with multi-label classification problems.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。