| 학회 |
한국공업화학회 |
| 학술대회 |
2022년 봄 (05/11 ~ 05/13, 제주국제컨벤션센터(ICC JEJU)) |
| 권호 |
26권 1호 |
| 발표분야 |
[콜로이드·계면화학] 피부 바이오 혁신연구동향 |
| 제목 |
Sensory predictive modeling of cosmetic formulations with rheology and machine learning |
| 초록 |
We suggest a novel predictive model for thespreadability of cosmetic formulations via LAOS analysis and machine learningtechniques. Rheological measurements of cosmetics formulations including thetransient elastic and viscous modulus from the sequence of physical process (SPP)analysis are selected as features for the three-variable predictive models, andthe spreadability of each formulation, which is quantitatively rated by trainedpanels, is set up as the target variable. Simple linear prediction models are derivedby the gradient descent algorithm, and it is shown that models based on the LAOSanalysis have better performance over models based on linear viscoelasticity. Additionally,non-linear prediction model is built based on the random forest algorithm withconsideration for nonlinear correlation between rheological measurements andspreadability. As is the case of the linear prediction model, the LAOS-basedprediction model shows better performance. |
| 저자 |
박준동1, 이수현1, 김성렬2, 이효정1, 오희묵3, 이준배3, 박경혜3, 이윤주3, 박천호3
|
| 소속 |
1숙명여자대, 2금오공과대, 3코스맥스 |
| 키워드 |
Cosmetics; Rheology; Spreadability; Sensory evaluation
|
| E-Mail |
|
