<p>The rapid expansion of biomedical datasets generated by international research programs across Africa highlights an urgent need for advanced data science expertise to translate these resources into impactful interventions. We designed an innovative training model for a data science research program funded through the National Institutes of Health Data Science for Health Discovery and Innovation in Africa initiative. The two-phase training model combined foundational instruction, integrating machine learning with traditional statistical methods, followed by intensive group- and competition-based learning. The model was tailored to leverage and analyze multimodal data warehouses from completed research projects. The inaugural datathon united trainees from 14 African countries, leveraging data sources for a large cohort study on malaria. Our pedagogical strategy bridged traditional statistical methods with analog machine learning approaches, illustrated through case studies on regression and image analysis. Each datathon team initiated a research project that culminates in a scientific manuscript. This paper details our complete datathon learning model, including its core competencies, learning objectives, and evaluation metrics, offering a comprehensive resource for researchers seeking to implement similar programs. The datathon framework provides a practical platform for advancing data science skills, fostering multidisciplinary research, and maximizing the impact of biomedical data resources.</p>

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

Advancing data science research education in Africa through datathon-driven innovations

  • Seydou Doumbia,
  • Fousseyni Kane,
  • Oudou Diabate,
  • Cheickna Cisse,
  • Ibrahim Sanogo,
  • Mamadou D. Coulibaly,
  • Fatoumata G. Fofana,
  • Alexandre Delamou,
  • Abdoul Habib Beavogui,
  • Sidibé M’Baye Thiam,
  • Jian Li,
  • Moussa Keïta,
  • Nafomon Sogoba,
  • Cheick Oumar Tangara,
  • Samuel Kakraba,
  • Mamadou Wele,
  • Mahamadou Diakite,
  • Mahamoudou Toure,
  • Jeffrey G. Shaffer

摘要

The rapid expansion of biomedical datasets generated by international research programs across Africa highlights an urgent need for advanced data science expertise to translate these resources into impactful interventions. We designed an innovative training model for a data science research program funded through the National Institutes of Health Data Science for Health Discovery and Innovation in Africa initiative. The two-phase training model combined foundational instruction, integrating machine learning with traditional statistical methods, followed by intensive group- and competition-based learning. The model was tailored to leverage and analyze multimodal data warehouses from completed research projects. The inaugural datathon united trainees from 14 African countries, leveraging data sources for a large cohort study on malaria. Our pedagogical strategy bridged traditional statistical methods with analog machine learning approaches, illustrated through case studies on regression and image analysis. Each datathon team initiated a research project that culminates in a scientific manuscript. This paper details our complete datathon learning model, including its core competencies, learning objectives, and evaluation metrics, offering a comprehensive resource for researchers seeking to implement similar programs. The datathon framework provides a practical platform for advancing data science skills, fostering multidisciplinary research, and maximizing the impact of biomedical data resources.