BROWSE
Detail
Conference Paper
시계열 제조 데이터 기반 AI 모델 개발을 위한 에이전트 플랫폼 설계 및 구현
Cited - time in 
Share 
Share
- Issue Date
- 2026-06
- Citation
- 대한전자공학회 학술 대회 (하계) 2026, pp.1-3
- Publisher
- 대한전자공학회
- Language
- Korean
- Type
- Conference Paper
- Abstract
- This paper proposes a rule‑based agent platform architecture to support the development of quality inspection AI models for manufacturing enterprises. The platform automates key processes in AI model development, including data processing, model selection, training strategy determination, performance evaluation, and inference service integration, through a layered structure comprising user input, agent‑based control, time‑series model pipeline, transfer learning model, and model storage and service. Implementation results show that candidate backbone model validation and selection based on user data enable automatic identification of suitable models, thereby streamlining AI model development with time‑series manufacturing data.
- KSP Keywords
- Agent Platform, Backbone Model, Data processing, Layered structure, Manufacturing data, Model Validation, Performance evaluation, Series model, Transfer learning, User data, User input