Back propagation is employed to forecast the current of a storm with various characteristics of storm surge; the technique is thus important in disaster forecasting. One of the most fuzzy types of information in the p...Back propagation is employed to forecast the current of a storm with various characteristics of storm surge; the technique is thus important in disaster forecasting. One of the most fuzzy types of information in the prediction of geological calamity is handled employing the information diffusion method. First, a single-step prediction model and neural network prediction model are employed to collect influential information used to predict the extreme tide level. Second, information is obtained using the information diffusion method, which improves the precision of risk recognition when there is insufficient information. Experiments demonstrate that the method proposed in this paper is simple and effective and provides better forecast results than other methods. Future work will focus on a more precise forecast model.展开更多
Aiming at the selection of fuzzy AHP and fuzzy DH methods in the previous studies, this paper evaluate the qualitative index system using expert questionnaire, the self-learning BP neural network model to construct th...Aiming at the selection of fuzzy AHP and fuzzy DH methods in the previous studies, this paper evaluate the qualitative index system using expert questionnaire, the self-learning BP neural network model to construct the index of system, and complete the establishment of model, in order to avoid the serious subjectivity, and using statistical and measurement methods test the reliability index, analyze the validity of the evaluation index system and completeness. Finally, the paper validate the practicability of the model.展开更多
基金Supported by the MISSION 908 (Nos. 908-02-03-07, SD-908-02-08)
文摘Back propagation is employed to forecast the current of a storm with various characteristics of storm surge; the technique is thus important in disaster forecasting. One of the most fuzzy types of information in the prediction of geological calamity is handled employing the information diffusion method. First, a single-step prediction model and neural network prediction model are employed to collect influential information used to predict the extreme tide level. Second, information is obtained using the information diffusion method, which improves the precision of risk recognition when there is insufficient information. Experiments demonstrate that the method proposed in this paper is simple and effective and provides better forecast results than other methods. Future work will focus on a more precise forecast model.
文摘Aiming at the selection of fuzzy AHP and fuzzy DH methods in the previous studies, this paper evaluate the qualitative index system using expert questionnaire, the self-learning BP neural network model to construct the index of system, and complete the establishment of model, in order to avoid the serious subjectivity, and using statistical and measurement methods test the reliability index, analyze the validity of the evaluation index system and completeness. Finally, the paper validate the practicability of the model.