The sparsely-activated models have achieved great success in natural language processing through large-scale parameters and relatively low computational cost, and gradually become a feasible technique for training and...
详细信息
With the exponential growth of mobile traffic data, mobile traffic classification is in a great need. It is an essential step to improve the performance of network services such as QoS and security monitoring. However...
详细信息
ISBN:
(数字)9781728186955
ISBN:
(纸本)9781728186962
With the exponential growth of mobile traffic data, mobile traffic classification is in a great need. It is an essential step to improve the performance of network services such as QoS and security monitoring. However, the widespread use of encrypted protocols, especially the TLS protocol, has posed great challenges to traditional traffic classification techniques. As the rule-based deep packet inspection approaches are ineffective for encrypted traffic classification, various machine learning methods have been studied and used. Recently, deep learning solutions which enable automatic feature extraction are also proposed to classify encrypted traffic. In this paper we propose App-Net, an end-to-end hybrid neural network, to learn effective features from raw TLS flows for mobile app identification. App-Net is designed by combining RNN and CNN in a parallel way. So that it can learn a joint flow-app embedding to characterize both flow sequence patterns and unique app signatures. We evaluate App-Net on a real-world dataset that covering 80 apps. The results show that our method can achieve an excellent performance and outperform the state-of-the-art methods.
Aiming at the fine-grained sentiment classification that distinguishes the emotional intensity, the commonly used dataset SST-1 is analyzed in depth. Through the analysis, it is found that the dataset has serious prob...
详细信息
With the rapid development of Internet technology, various network attack methods come out one after the other. SQL injection has become one of the most severe threats to Web applications and seriously threatens vario...
详细信息
With the rapid development of Internet technology, various network attack methods come out one after the other. SQL injection has become one of the most severe threats to Web applications and seriously threatens various Web application services and users' data security. There are both traditional detection methods and emerging methods based on deep learning technology with higher detection accuracy for the detection of SQL injection. However, they are all for detecting a single statement and cannot determine the stage of the attack. To further improve the effect of SQL injection detection, this paper proposes an integrated detection framework for SQL injection behavior based on both text features and traffic features. We propose a SQL-LSTM model based on deep learning technology as the detection model at the text features level. Meanwhile,the features of the data traffic are merged. By this integrated method, the detection effect of SQL injection is further improved.
Convolutional layers are ubiquitous in a variety of deep neural networks. Due to the lower computation complexity and the smaller number of parameters, convolutions with small filter sizes are often used, such as one ...
详细信息
ISBN:
(数字)9781728169811
ISBN:
(纸本)9781728169828
Convolutional layers are ubiquitous in a variety of deep neural networks. Due to the lower computation complexity and the smaller number of parameters, convolutions with small filter sizes are often used, such as one by one convolution. Nevertheless, these small convolution operations are still time-consuming. A common approach to implementing convolutions is to transform them into matrix multiplications, known as GEMM-based convolutions. The approach maybe incurs additional memory overhead and calls matrix multiplication routines, which are not optimized for matrices generated by convolutions. In this paper, we present a new parallel one by one direct convolution implementation on ARMv8 multi-core CPUs, which doesn't incur any additional memory space requirement. Our implementation is verified on two ARMv8 CPUs, Phytium FT-1500A and FT-2000plus. In terms of performance and scalability, our implementation is better than GEMM-based implementations in all the tests on Phytium FT-1500A. On Phytium FT-2000plus, our approach gives much better performance and scalability than GEMM-based approaches in most cases.
The following topics are dealt with: learning (artificial intelligence); text analysis; natural language processing; pattern classification; graph theory; data mining; security of data; social networking (online); fea...
ISBN:
(数字)9781728195582
ISBN:
(纸本)9781728195599
The following topics are dealt with: learning (artificial intelligence); text analysis; natural language processing; pattern classification; graph theory; data mining; security of data; social networking (online); feature extraction; and neural nets.
Static random access memory (SRAM) on field programmable gate arrays (FPGAs) can be emulated to offer ternary content addressable memory (TCAM) functionality. However, SRAM-based TCAM wastes storage resources. This is...
详细信息
Static random access memory (SRAM) on field programmable gate arrays (FPGAs) can be emulated to offer ternary content addressable memory (TCAM) functionality. However, SRAM-based TCAM wastes storage resources. This is due to the limited capacity of the physical addresses in the SRAM unit. This work proposes a LUTRAM-based TACM scheme on the FPGA called Memory-Efficient TCAM (ME-TCAM). METCAM divides SRAM unit into multiple virtual blocks mapping to a portion of the TCAM table to store the more address information of the TCAM table. Operation on SRAM block means that increasing the overall emulated TCAM bits/SRAM. Moreover, ME-TCAM exploits Xilinx primitives to conFigure lookup tables (LUTs) as 32 × 2 lookup table RAMs (LUTRAMs). We implement ME-TCAM using LUTRAM with a size of 512 × 48 and 1024 × 144 on a Virtex-7 FPGA device. Compared with the state-of-the-art research DUR, ME-TCAM achieves at least 2.6 times more memory efficiency.
Dear editor,Docker1), as a de-facto industry standard [1], enables the packaging of an application with all its dependencies and execution environment in a light-weight, self-contained unit, i.e., *** launching the co...
详细信息
Dear editor,Docker1), as a de-facto industry standard [1], enables the packaging of an application with all its dependencies and execution environment in a light-weight, self-contained unit, i.e., *** launching the container from Docker image, developers can easily share the same operating system, libraries, and binaries [2]. As the configuration file, the dockerfile plays an important role,
The hippocampus plays a vital role in the diagnosis and treatment of many neurological disorders. Recent years, deep learning technology has made great progress in the field of medical image segmentation, and the perf...
详细信息
Knowledge representation learning (KRL) is one of the important research topics in artificial intelligence and Natural language processing. It can efficiently calculate the semantics of entities and relations in a low...
ISBN:
(数字)9781728195582
ISBN:
(纸本)9781728195599
Knowledge representation learning (KRL) is one of the important research topics in artificial intelligence and Natural language processing. It can efficiently calculate the semantics of entities and relations in a low-dimensional space, and effectively solve the problem of data sparsity, which significantly improve the performance of knowledge acquisition, fusion and reasoning and so on. Starting from the five perspectives of distance-based, semantic matching, bilinear-based, neural network model and additional information model, this paper first introduces the overall framework and specific model design, and then correspondingly introduces the experimental evaluation tasks, metrics and benchmark datasets of each model. On this basis, how to apply KRL to various downstream tasks is summarized.
暂无评论