Scholarlink

This paper presents GridLife, a non-parametric grid-based clustering framework inspired by Conway’s Game of Life. It maps data densities to a discrete lattice and evolves the structures using shape-preserving cellular automata dynamics using core and dense-dead regions. It shows density-based clustering behavior with near-linear scalability under fixed grid resolution. Experiments on synthetic datasets provide comparable performance to canonical clustering methods, including non-linearly separable cases, with favorable computational efficiency.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

GridLife: A Game of Life Inspired Non-parametric Grid-Based Linear-Scalable Density Evolution Framework for Clustering

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Jit Mukherjee

摘要

This paper presents GridLife, a non-parametric grid-based clustering framework inspired by Conway’s Game of Life. It maps data densities to a discrete lattice and evolves the structures using shape-preserving cellular automata dynamics using core and dense-dead regions. It shows density-based clustering behavior with near-linear scalability under fixed grid resolution. Experiments on synthetic datasets provide comparable performance to canonical clustering methods, including non-linearly separable cases, with favorable computational efficiency.