Scholarlink

VIS4AI presents many research challenges and opportunities across the entire machine learning lifecycle. At every stage, from data preparation to model development and model deployment, researchers and practitioners face unique hurdles that demand innovative visualization solutions to ensure efficient and effective AI systems. Beyond these stage-specific challenges, generic research challenges span the entire machine learning lifecycle. Furthermore, the prevalence of foundation models in recent years has added complexity to VIS4AI [268]. This demands even more innovative visualization methods to improve and adapt these models. In this chapter, we explore key challenges and potential avenues for further exploration.

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

Research Challenges and Opportunities

  • 章节

  • 发表日期：

    2025年

查看全文
  • Shixia Liu,
  • Weikai Yang,
  • Junpeng Wang,
  • Jun Yuan

摘要

VIS4AI presents many research challenges and opportunities across the entire machine learning lifecycle. At every stage, from data preparation to model development and model deployment, researchers and practitioners face unique hurdles that demand innovative visualization solutions to ensure efficient and effective AI systems. Beyond these stage-specific challenges, generic research challenges span the entire machine learning lifecycle. Furthermore, the prevalence of foundation models in recent years has added complexity to VIS4AI [268]. This demands even more innovative visualization methods to improve and adapt these models. In this chapter, we explore key challenges and potential avenues for further exploration.