Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2020-09-28 Cooperative journals: 《计算机应用研究》
Abstract： Software Defined Networking (SDN) has attracted much attention in recent years. Some scholars have integrated it with smart city construction. In view of the current situation of decentralized traffic management and combined with the characteristics of SDN forwarding and control separation and centralized control, introduces it into Intelligent Transport System (ITS) , and proposes a collaborative management architecture to achieve collaborative management and improve traffic. Firstly, the characteristics of SDN are expounded, and an intelligent traffic architecture based on SDN is proposed by using ITS characteristics and combining with ITS. Secondly, a collaborative management strategy model is built combining factors such as dynamic speed limit, path guidance and weather and compared with a single control strategy. Finally, through simulation experiments, the proposed architecture strategy was verified and analyzed. The experimental results show that compared with the single control strategy, the management under this architecture smooths the traffic flow, reduces the average delay at the intersection by 20.7%, reduces the average queue length by 35%.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-05-10 Cooperative journals: 《计算机应用研究》
Abstract： In a blockchain network based on Proof of work(Pow) , miners usually choose to join the mining pool. As there are many mining pools and different mining pools have different computing power and may adopt different reward mechanisms, miners can get different rewards in different mining pools. For the choice of mining pool faced by the miners, This paper proposed a mining pool selection model based on risk decision criteria, aiming at the problem of miners' selection of mining pools, and studied the effect of computing power and reward mechanism on the miners’ optimal pool selection decisions. Firstly, calculated the miners’ reward and given the reward matrix, Secondly, by using the maximum likelihood criterion and the expectation criterion respectively derived the optimal selection strategy, Finally, through simulation experiments, validated the proposed pool selection strategies. Experimental results show that compared with the simple strategy, the proposed strategy can bring higher rewards to the miners in most cases.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-01-28 Cooperative journals: 《计算机应用研究》
Abstract： In recent years, open source software has frequently exposed high-risk vulnerabilities, posing a huge threat to the security of enterprise information system. Aiming at the open source software vulnerability, this paper proposed a software source code vulnerability detection method based on deep clustering algorithm. This method uses code graph model to construct the code attribute map and traverses the key code nodes to extract the application programming interfaces (API) sequence, then takes the key sequence as the center to cluster and calculates the outliers of the function in each clustering to generate a test report, matches the vulnerability library to detect vulnerabilities in the source code. The experimental results show that the proposed method can locate the key code segments of the vulnerability in open source software and detect the vulnerability.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-09-12 Cooperative journals: 《计算机应用研究》
Abstract： Aiming at the fact that partial occlusion affects facial expression recognition in practical applications, an expression recognition method based on generative adversarial networks (GAN) is proposed. Firstly, the occlusion face images are filled and repaired, and then the expression recognition is performed. The generator of GAN is composed of a convolutional Auto-encoder, the face images generated by adversarial learning between generator and discriminator are more vivid. The discriminator is composed of the convolutional neural network and it has good feature extraction ability, and a multi-classification layer is added to construct the expression classifier, which avoids feature re-calculation. In order to solve the problem of insufficient training samples, the celebA face dataset is used to train face filling and repairing, and the feature extraction part of the expression classifier is pre-trained. Experiments on the CK+ dataset show that the face images after filling are real and coherent, and a higher expression recognition rate is achieved, especially the recognition rate of large-area occlusion of the face is improved.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-19 Cooperative journals: 《计算机应用研究》
Abstract： In the current aspect sentiment analysis task, the traditional attention-based deep learning model lacks the effective attention to the aspect information and the sentiment information. This paper put forward a new LSTM model, which combining aspect information and deeper attention. Through the bidirectional LSTM with shared weights, it trained the aspect embedding and the text embedding to get the aspect feature and text feature to carry on the feature fusion, and after the deeper attention mechanism processing, it obtained the classification result of the corresponding aspect by the classifier. The experimental results of the SemEval-2014 Task4 and SemEval-2017 Task4 datasets show that this method has further improved the accuracy and stability of the attention-based sentiment analysis model in the aspect sentiment analysis. The introduction of aspect features and deeper attention mechanisms is of great significance to the task of sentiment analysis based on aspect, which provides method support for public opinion analysis, question answering system and text reasoning.
Hits
Hits
Downloads
Comment