Presentations · 2020.10.13
AI training-data quality checks and maintenance: NIA’s 2020 presentation
인공지능 데이터 품질 및 이슈
Overview
This 2020 presentation distinguishes raw-data suitability, labeling accuracy and machine-learning validity. With examples including autonomous-driving data, it discusses classification rules, dataset splits, domain-expert involvement and continuing maintenance during data use.
Publication and version information
The date is the presentation day in the event programme; the PDF posting date is unverified.