[论文解读] Semi-Supervised Constrained Clustering: An In-Depth Overview, Ranked Taxonomy and Future Research Directions

Germán González-Almagro, Daniel Peralta+3 more|arXiv (Cornell University)|Feb 28, 2023

Advanced Clustering Algorithms Research被引用 10

一句话总结

本综述提供了对约束聚类方法的综合分类与排序，分析背景知识类型，并概述未来研究方向。

ABSTRACT

Clustering is a well-known unsupervised machine learning approach capable of automatically grouping discrete sets of instances with similar characteristics. Constrained clustering is a semi-supervised extension to this process that can be used when expert knowledge is available to indicate constraints that can be exploited. Well-known examples of such constraints are must-link (indicating that two instances belong to the same group) and cannot-link (two instances definitely do not belong together). The research area of constrained clustering has grown significantly over the years with a large variety of new algorithms and more advanced types of constraints being proposed. However, no unifying overview is available to easily understand the wide variety of available methods, constraints and benchmarks. To remedy this, this study presents in-detail the background of constrained clustering and provides a novel ranked taxonomy of the types of constraints that can be used in constrained clustering. In addition, it focuses on the instance-level pairwise constraints, and gives an overview of its applications and its historical context. Finally, it presents a statistical analysis covering 307 constrained clustering methods, categorizes them according to their features, and provides a ranking score indicating which methods have the most potential based on their popularity and validation quality. Finally, based upon this analysis, potential pitfalls and future research directions are provided.

研究动机与目标

• 提供关于约束聚类及背景知识类型的详细背景介绍。
• 给出在约束聚类中使用的约束类型的新型有序分类法。
• 综述实例层面的成对约束及其应用与历史。
• 进行对约束聚类方法的统计分析，以按流行度和验证质量进行排序。
• 提出约束聚类中的潜在陷阱及未来研究方向。

提出的方法

• 引入半监督聚类中使用的背景知识分类法（分区级、实例级、聚类级、特征级、距离级；以及杂项）。
• 形式化约束聚类及成对约束（必须链接、不能链接）及其扩展（可能链接、模糊、精英、排序）。
• 分析部分聚类和层次聚类设定下约束聚类的可行性与复杂性，包括层次 CC 的死胡同。
• 综述约束聚类方法的历史发展、应用以及广泛的约束聚类方法语料库（统计抽样与排序）。
• 提出一个评分/排序系统，以按特征和验证质量对 307 种约束聚类方法进行评估和排序。

实验结果

研究问题

• RQ1 半监督约束聚类中使用的背景知识类型有哪些，它们可以如何分类？
• RQ2 成对约束与其他形式的约束如何影响约束聚类的可行性、复杂性和实际性能？
• RQ3 根据所提出的排序系统，哪些约束聚类方法最具影响力或最有前景？
• RQ4 约束聚类研究中的常见陷阱与未来方向是什么？

主要发现

• 提供关于 CC 的约束类型和背景知识的全面分类。
• 表明对 CL 约束的可行性问题对部分聚类与层次聚类的 CC 均为 NP-完全。
• 识别包括成对、分组、三元组以及与层次相关的约束在内的大范围约束形式。
• 提出一个评分系统和排序分类法，按流行度和验证质量对 307 种方法进行评估。
• 突出潜在陷阱并概述约束聚类的未来研究方向。