Scholarlink

This article introduces a novel methodology for performing univariate and multivariate functional clustering analysis (FCA) on sparse data. The approach aims to address data sparsity by reconstructing missing functional data using the Conditional Expectation (PACE) method. The FCA algorithm uses the EM (Expectation-Maximization) method and incorporates the Bayesian Information Criterion (BIC) for model validation. An application on real functional data is also shown.

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

Functional Sparse Data Clustering Using Conditional Expectation PACE Method

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Si-Ahmed Idris,
  • Hamdad Leila,
  • Dabo-Niang Sophie

摘要

This article introduces a novel methodology for performing univariate and multivariate functional clustering analysis (FCA) on sparse data. The approach aims to address data sparsity by reconstructing missing functional data using the Conditional Expectation (PACE) method. The FCA algorithm uses the EM (Expectation-Maximization) method and incorporates the Bayesian Information Criterion (BIC) for model validation. An application on real functional data is also shown.