Scholarlink

This paper presents DataLens, a comprehensive platform designed to enhance data-driven decision-making through automated preprocessing and insightful data analysis. As the volume and complexity of data increase, traditional Exploratory Data Analysis (EDA) methods: characterized by manual steps, become increasingly inefficient and error-prone. DataLens addresses these challenges with its AutoPreprocessor, achieving a competitive accuracy of 84.5%, a low inference time of 1.8 s, and a minimal memory footprint of 180 MB, positioning it favorably against established frameworks such as Google AutoML and H2O AutoML. By automating the preprocessing and reporting processes, DataLens significantly streamlines EDA, allowing analysts to focus on model building and strategic decision-making. The platform also includes the Bird Eye feature for comprehensive data visualization and the Insights Generator for transforming complex analytical results into actionable insights. Through a comparative analysis of various automated preprocessing frameworks, this study demonstrates DataLens’s potential to optimize workflows and enhance productivity across diverse industries. Ultimately, DataLens aims to redefine data exploration and empower users with high-quality outputs.

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

Automating Data Discovery: Bridging EDA and AutoML for Smarter, Faster Insights

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Prutha Annadate,
  • Neha Aher,
  • Suhrud Joshi,
  • Rishabh Gupta,
  • Nitin Pise

摘要

This paper presents DataLens, a comprehensive platform designed to enhance data-driven decision-making through automated preprocessing and insightful data analysis. As the volume and complexity of data increase, traditional Exploratory Data Analysis (EDA) methods: characterized by manual steps, become increasingly inefficient and error-prone. DataLens addresses these challenges with its AutoPreprocessor, achieving a competitive accuracy of 84.5%, a low inference time of 1.8 s, and a minimal memory footprint of 180 MB, positioning it favorably against established frameworks such as Google AutoML and H2O AutoML. By automating the preprocessing and reporting processes, DataLens significantly streamlines EDA, allowing analysts to focus on model building and strategic decision-making. The platform also includes the Bird Eye feature for comprehensive data visualization and the Insights Generator for transforming complex analytical results into actionable insights. Through a comparative analysis of various automated preprocessing frameworks, this study demonstrates DataLens’s potential to optimize workflows and enhance productivity across diverse industries. Ultimately, DataLens aims to redefine data exploration and empower users with high-quality outputs.