Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-01-03 Cooperative journals: 《计算机应用研究》
Abstract： Water quality prediction and eutrophication analysis of lakes are important technical means of water pollution prevention and control. However, the existing water quality prediction research is usually in a form of single-valued prediction, and analyze eutrophication status on this basis, which has a certain degree of haphazard and uncertainty. Combining with the water quality kinetic model, proposed a water quality prediction and eutrophication risk assessment method based on Monte-Carlo simulation. Based on the prior distribution of water quality index and model parameters of water quality kinetic model, used Monte Carlo simulation to predict the evolution of water quality index to obtain the probability distribution of water quality indicators in future time and achieve water quality prediction. Further, constructed an integrated eutrophication status index. Combining with the predicted results of water quality indexes, calculated the probability distribution of comprehensive nutritional status index and the probability of different nutritional status to assess the eutrophication risk. The simulation results show that the proposed method can effectively predict the water quality and analyze eutrophication status, with more comprehensive consideration and accuracy. Meanwhile, it overcomes the haphazard brought by single-valued prediction result.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-01-03 Cooperative journals: 《计算机应用研究》
Abstract： Passwords are universal methods for data encryption and user authentication. The passwords set by users are not completely random. Therefore, the passwords are easily guessed and attacked by password crackers. Using a password guessing algorithm is an effective way to assess the strength and security of a password. This paper proposes PassCVAE based on conditional variation auto-encoding (CVAE) model. The algorithm take user's personal information as the conditional feature to train the password attack model. For the encoder, bidirectional GRU recurrent neural network and Text Convolution Neural Network (TextCNN) are used to extract the feature of the password sequence and personal information Abstract: y. The decoder uses two layers of GRU neural network to generate a password sequence based on the corresponding feature of personal information and hidden coding of password. The algorithm can effectively fit the distribution of password data, learn character combination rules and generate high-quality password guessing data. Multiple sets of experiments show that the proposed PassCVAE is better than the existing password guessing algorithms.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-08-13 Cooperative journals: 《计算机应用研究》
Abstract： In view of the low resource of endangered languages, the establishment and research of end-to-end speech recognition model can explore new ways for the protection and transmission of endangered languages. this paper combined dynamic bi-directional long short-term memory network and connectionist temporal classification model into an end-to-end speech recognition model. When performing phoneme-level recognition training, the batch size of the data passed into the model can be adaptively adjusted according to the training model, which not only speeds up the convergence but also improves the generalization of the model. By adjusting the hierarchy of the deep neural network and extracting different phonetic features for model comparison, the experimental results show that both the endangered languages - Lvsu and Tujia have good recognition results.
Hits
Hits
Downloads
Comment