| 학회 |
한국화학공학회 |
| 학술대회 |
2017년 가을 (10/25 ~ 10/27, 대전컨벤션센터) |
| 권호 |
23권 2호, p.1585 |
| 발표분야 |
공정시스템 |
| 제목 |
PCB 물리화학변수 추정을 위한 Bayesian Regularized Neural Network 기반 QSPR 모델개발 |
| 초록 |
The aim of any eco-toxicological is to determine the efficiency of the developed QSAR model for the toxicity and physiochemical properties of chemical compounds. Polychlorinated biphenyls (PCBs) are capable of causing a large range of adverse effects. This research proposes a Bayesian approach to developed new quantitative structure-property relationship model to predict the physiochemical properties of PCBs. The prediction accuracy of proposed method was investigated and compared the results from OECD QSAR toolbox. The coefficient of determination (R2) and sum of square error (PRESS) were utilized to check the performance of the proposed model i.e. 0.9714 and 2.4286 respectively. The proposed QSAR model can capture the predictive features in determining the activities of chemicals. Acknowledgments: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIP) (No.2015R1A2A2A11001120). |
| 저자 |
남기전, Usman Safder, 유창규
|
| 소속 |
경희대 |
| 키워드 |
화학 및 생물공정; 공정모델링; 공정모사
|
| E-Mail |
|
| 원문파일 |
초록 보기 |
