ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 Weak False Label Learning Model for Sensor Data Recognition
Cited 0 time in scopus Download 0 time Share share facebook twitter linkedin kakaostory
저자
장성준, 이헌주
발행일
201510
출처
International Conference on Control, Automation and Systems (ICCAS) 2015, pp.1321-1323
DOI
https://dx.doi.org/10.1109/ICCAS.2015.7364842
협약과제
15MS5700, 후각 바이오 정보 기반 감성증강 인터랙티브 콘텐츠 기술 개발, 이해룡
초록
Real world behavior recognitions tend to suffer from incomplete data because sensors are not perfect. Although machine learning algorithms are successfully applied to recognitions, they do not work well in multi-valued output functions because true and false label in same input collide in learning process. In this paper, we propose a noble algorithm which lessens multi-valued function's problem by weakening false labels. It also includes virtual samples and output normalization to compensate for the balance between true and false labels.
키워드
Accelerometer, Artificial Neural Network, Human Behavior Recognition, Sensor
KSP 제안 키워드
Artificial neural networks, Data recognition, Human behavior recognition, Learning model, Machine Learning Algorithms, Multi-valued function, Real-world, incomplete data, learning process, sensor data, virtual samples