| 학회 | 한국화학공학회 |
| 학술대회 | 2006년 가을 (10/27 ~ 10/28, 고려대학교) |
| 권호 | 12권 2호, p.1493 |
| 발표분야 | 공정시스템 |
| 제목 | Data mining software development for high-throughput screening of heterogeneous catalysts |
| 초록 | Large amounts of experimental data are simultaneously accumulated in a relatively short period of time by High-Throughput Experimentation (HTE). The large number of variables in play and the applications of complex optimization algorithms for the experimental design make the direct human interpretation of data derived from HTE very difficult. Artificial neural networks using multi-layer perceptrons have been successfully applied to modeling for new catalyst development. But the complexity of the network significantly affects its learning capability and generalization. So we propose a new self-organizing algorithm of artificial neural networks that always matches the complexity of the model to that of the problem. Then the algorithm is applied to the data mining software for high-throughput screening of heterogeneous catalysts. Acknowledgement: This work is supported by Center for Ultramicrochemical Process Systems sponsored by KOSEF. |
| 저자 | 강수길, 박선원 |
| 소속 | 한국과학기술원 |
| 키워드 | high-throughput screening; artificial neural networks; data mining; catalysts |
| 원문파일 | 초록 보기 |
