Considering the growing volume of available information, storing and managing data is an increasing challenge. Leveraging these raw datasets to extract meaningful knowledge and make informed decisions requires data mining methods, especially when combined with Machine Learning algorithms to uncover hidden patterns in data. Considering this context, this work describes a project to contribute to data-driven initiatives by seamlessly integrating a data repository system with Machine Learning (ML) tools. This integration allows users to implement pipelines within a single system—from storing and managing datasets to training and deploying ML models, making easier, more efficient, and intuitive data science workflow accessible to all, not just computer scientists. Furthermore, the system aims to enhance knowledge sharing among researchers and practitioners by enabling users to share datasets and trained machine learning models within the platform. This work highlights the system’s key features, particularly the Machine Learning module, which includes classic ML algorithms for tabular data and the newly introduced Deep Learning algorithms for image classification. The new Model Sharing module also emphasizes promoting collaboration by allowing users to share and manage trained models effectively.

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

Exploring Data Symbion EI Deep Learning and Model Sharing Modules

  • Rafael Huszcza,
  • Amanda Mendes,
  • Jeferson Lopes,
  • Eduardo N. Borges,
  • Giancarlo Lucca,
  • Pablo D. B. Guilherme,
  • Leandro A. Pereira

摘要

Considering the growing volume of available information, storing and managing data is an increasing challenge. Leveraging these raw datasets to extract meaningful knowledge and make informed decisions requires data mining methods, especially when combined with Machine Learning algorithms to uncover hidden patterns in data. Considering this context, this work describes a project to contribute to data-driven initiatives by seamlessly integrating a data repository system with Machine Learning (ML) tools. This integration allows users to implement pipelines within a single system—from storing and managing datasets to training and deploying ML models, making easier, more efficient, and intuitive data science workflow accessible to all, not just computer scientists. Furthermore, the system aims to enhance knowledge sharing among researchers and practitioners by enabling users to share datasets and trained machine learning models within the platform. This work highlights the system’s key features, particularly the Machine Learning module, which includes classic ML algorithms for tabular data and the newly introduced Deep Learning algorithms for image classification. The new Model Sharing module also emphasizes promoting collaboration by allowing users to share and manage trained models effectively.