研究人员为 k-means++ 算法开发了一种名为 GTRC 的新准则，用于确定最佳重启次数。该方法使用 Good-Turing 估计和置信区间，根据数据集难度动态调整重启次数，而不是依赖于任意固定的次数。在 36 个数据集上的测试表明，GTRC 在适当变化重启次数的同时实现了具有竞争力的聚类质量，提供了一种更具原则性的方法。 AI
影响 为优化聚类算法提供了一种更具原则性和可解释性的方法，有望提高机器学习任务的效率和结果。
排序理由 该集群包含一篇关于 k-means++ 新算法准则的学术论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
The k-means++ algorithm is commonly restarted multiple times to avoid poor local optima, yet the number of restarts is almost always chosen arbitrarily and applied uniformly regardless of data set difficulty. This undermines any comparison relying on such a choice and wastes comp…
arXiv:2607.08243v1 Announce Type: cross Abstract: The k-means++ algorithm is commonly restarted multiple times to avoid poor local optima, yet the number of restarts is almost always chosen arbitrarily and applied uniformly regardless of data set difficulty. This undermines any c…
The k-means++ algorithm is commonly restarted multiple times to avoid poor local optima, yet the number of restarts is almost always chosen arbitrarily and applied uniformly regardless of data set difficulty. This undermines any comparison relying on such a choice and wastes comp…