Enhancing the Effectiveness of Fungicides by Optimizing Their Combinations
In controlling biological diseases,it is often more potent to use a combination of agents than using individual ones.However,the number of possible combinations increases exponentially with the number of agents and their concentrations.It is prohibitive to search for effective agent combinations by trial and error as biological systems are complex and their responses to agents are often a slow process.This motivates to build a suitable model to describe the biological systems and help reduce the number of experiments.In this paper,we consider the use of fungicides to inhibit Bipolaris maydis and construct models that describe the responses to fungicide combinations.Three data-driven modeling methods,the polynomial regression,the artificial neural network and the support vector regression,are compared based on the experimental data of the inhibition rates of the southern corn leaf blight with different fungicide combinations.The analysis of the results demonstrates that the support vector regression is best suited to the construction of the response model in terms of achieving better prediction with fewer experiments.
agent combinations complex system data-driven model polynomial regression artificial neural network support vector regression
Xiang Wang Jia Ma Xiaowei Li Xiaodong Zhao Zongli Lin Jie Chen Zhifeng Shao
Department of Automation,and Key Laboratory of System Control and Information Processing of the Mini Department of Resource and Environmental Science,School of Agriculture and Biology,Shanghai Jiao Ton School of Biomedical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China Charles L.Brown Department of Electrical and Computer Engineering,University of Virginia,P.O.Box 400
国际会议
长沙
英文
4666-4670
2014-05-31（万方平台首次上网日期，不代表论文的发表时间）