| 학회 | 한국공업화학회 |
| 학술대회 | 2022년 봄 (05/11 ~ 05/13, 제주국제컨벤션센터(ICC JEJU)) |
| 권호 | 26권 1호 |
| 발표분야 | [화학물질안전·위해성] 유해화학물질의 안전성과 위험성 평가 |
| 제목 | 머신러닝을 이용한 노비촉 유사물질 독성 및 물리화학적 특성 예측 연구 |
| 초록 | After recent terrorist attacks by Novichok agents and followingdecomposition operations, one knows that how important it is to know thephysicochemical properties such as vapor pressure and toxicity as well forunknown nerve agent structure. To prevent the continuous threat by new types ofnerve agents, we successfully developed ML models to predict one of the mostimportant physical property, vapor pressure to cope with the escape/removaloperation when it is leaked or used. Moreover, ML classification model issuccessfully built on organophosphorus compounds for toxicity prediction.Importantly, tuned ML model was successfully used for predicting the toxicityof novichok materials, which is described in Chemical Weapons Convention list. |
| 저자 | 정근홍 |
| 소속 | 육군사관학교 |
| 키워드 | 머신러닝; 양자계산; 화학작용제; 독성; 물리화학적 특성 |
