Scholarlink

This Chapter discusses several extensions and modifications of the core GP model to make it better suited for financial applications. Topics covered include heteroskedastic GPs (Sect. 3.1) to capture input-dependent noise and alternative likelihoods to tackle non-Gaussian noise (Sect. 3.2). We also survey multi-output GPs (Sect. 3.3) that provide a framework to capture correlation across multiple outputs, localization methods to combat nonstationarity or reduce runtime for large datasets (Sect. 3.4), and GP updating equations for streaming training data (Sect. 3.5).

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

Advanced GP Modeling Topics

  • 章节

  • 发表日期：

    2025年

查看全文
  • Michael Ludkovski,
  • Jimmy Risk

摘要

This Chapter discusses several extensions and modifications of the core GP model to make it better suited for financial applications. Topics covered include heteroskedastic GPs (Sect. 3.1) to capture input-dependent noise and alternative likelihoods to tackle non-Gaussian noise (Sect. 3.2). We also survey multi-output GPs (Sect. 3.3) that provide a framework to capture correlation across multiple outputs, localization methods to combat nonstationarity or reduce runtime for large datasets (Sect. 3.4), and GP updating equations for streaming training data (Sect. 3.5).