Browse > Article

A Study on Drift Phenomenon of Trained ML  

Shin, ByeongChun (전남대학교 수학과)
Cha, YoonSeok (제노테크(주))
Kim, Chaeyun (제노테크(주))
Cha, ByungRae (광주과학기술원 AI 대학원 & 제노테크(주))
Publication Information
Smart Media Journal / v.11, no.7, 2022 , pp. 61-69 More about this Journal
Abstract
In the learned machine learning, the performance of machine learning degrades at the same time as drift occurs in terms of learning models and learning data over time. As a solution to this problem, I would like to propose the concept and evaluation method of ML drift to determine the re-learning period of machine learning. An XAI test and an XAI test of an apple image were performed according to strawberry and clarity. In the case of strawberries, the change in the XAI analysis of ML models according to the clarity value was insignificant, and in the case of XAI of apple image, apples normally classified objects and heat map areas, but in the case of apple flowers and buds, the results were insignificant compared to strawberries and apples. This is expected to be caused by the lack of learning images of apple flowers and buds, and more apple flowers and buds will be studied and tested in the future.
Keywords
Machine Leaning; AI; Drift; IoU; XAI;
Citations & Related Records
연도 인용수 순위
  • Reference
1 IoU, https://deep-learning-study.tistory.com/402 (accessed Jul. 10, 2022)
2 Andrew P. McMahon, "Machine Learning Engineering with Python - Manage the production life cycle of machine learning models using MLOps with practical examples," Packt Publishing, 2021.
3 Frank Hutter, Lars Kotthoff, Joaquin Vanschroen, "Automated Machine Learning - Methods, Systems, Challenges," Springer, 2019.
4 NVIDIA KOREA, "설명 가능한 AI란 무엇인가?," https://blogs.nvidia.co.kr/2021/07/27/what-is-explainable-ai/ (accessed Jul. 11, 2022)
5 안재현, "XAI 설명 가능한 인공지능, 인공지능을 해부하다," 위키북스, 2020년
6 차윤석, 박진영, 박선, 김종원, 차병래, "ML 모델의 Drift 탐지를 위한 XAI 분석에 따른 머신러닝 모델 팩토리의 제안," 스마트미디어종합학술대회, 2022년 6월