In order to identify faults of power system measurement subsystem, a method based on deep belief network is proposed in this paper. Firstly, data from actual measurement system is collected and divided into training a...
详细信息
ISBN:
(纸本)9783030120825;9783030120818
In order to identify faults of power system measurement subsystem, a method based on deep belief network is proposed in this paper. Firstly, data from actual measurement system is collected and divided into training and test samples. And then, the data is used to train a deep belief network. Finally, the model's fault diagnosis results and actual samples' labels are combined as a cross-validation set to test the deep belief network. The results show that the method based on deep belief network proposed in this paper can be more stable and reliable identification of electric power measurement equipment fault diagnosis.
Information leakage by side channels in web applications is a major security threat, even when web traffic is encrypted. In this paper, we describe an automated system to analyse leakages of user privacy. Previous wor...
详细信息
ISBN:
(纸本)9781450387835
Information leakage by side channels in web applications is a major security threat, even when web traffic is encrypted. In this paper, we describe an automated system to analyse leakages of user privacy. Previous works focus on communications directly interacting with sensitive information. We show how, in real-world web applications, user information can be leaked through "non-intuitive" communications, which do not contain direct interactions with sensitive information. Unexpectedly, we found that user privacy can be leaked more from this kind of "non-intuitive" communication than direct interactions. This work also discloses user identities can be inferred by traffic analysis. In this work, we combine machine learning to recognize traffic pattern, i.e., Hidden Markov model is used to fingerprint web traffic of user privacy.
Convolutional Neural Network (CNN) is an efficient recognition algorithm used in patternrecognition, image processing, etc. The problem of how to build an efficient computational system for CNNs has become a pressing...
详细信息
Convolutional Neural Network (CNN) is an efficient recognition algorithm used in patternrecognition, image processing, etc. The problem of how to build an efficient computational system for CNNs has become a pressing one, which is traditionally achieved by software optimization or hardware accelerator design. In contrast, this paper argues that software algorithms and hardware architectures affect each other in neural network applications with a high degree of coupling. In this paper, we propose a software and hardware-based co-optimization design to improve CNNs’ processing efficiency by co-optimization of software and hardware. First, a modular analysis and collaborative design are carried out for the ShuffleNetV2 model by implementing quantization and improving the computational unit. Second, the model is optimized based on reconfigurable computing devices for its characteristics. Third, 8 bit quantization is implemented, while the depth-separable convolution operation and channel selection module is redesigned to make the module perform operations in a hardware-friendly form. The experimental work was carried out on the FPGA platform Xilin xzynqxc-7Z045 using High Level Synthesis (HLS). The experimental results show that the optimized model has a significant improvement in terms of resource utilization and latency.
In the age of Big Data, processing high volumes of data as fast as possible is an important task. Parallel computing is a useful tool that has become more and more available and widespread in the past decade, making i...
详细信息
ISBN:
(纸本)9781728144610;9781728144603
In the age of Big Data, processing high volumes of data as fast as possible is an important task. Parallel computing is a useful tool that has become more and more available and widespread in the past decade, making it possible to accelerate traditional data processing methods or create new ones that are built upon parallelization techniques. Sequential Fuzzy Indexing Tables are classifiers that expand the capabilities of Lookup Tables in order to achieve a fast classification. However, due to the size of their structure, they cannot be used for problems with larger complexity. In this paper, a new classifier architecture is proposed that uses parallelization techniques to be used in quick processing of large volumes of data.
Almost over 4 billion people are currently making rampant usage of the internet. The massive utilization of mobile technology along with the rise of the digital era caused a socio-technical threat to the Government an...
详细信息
Almost over 4 billion people are currently making rampant usage of the internet. The massive utilization of mobile technology along with the rise of the digital era caused a socio-technical threat to the Government and to the public. Many new developments in the internet and modern technologies give rise to new illegal and unethical opportunities among which some of them are crime. Cyber crime is an unlawful means which makes use of a digital media either as a tool or as a target or both. Cyber crime cases, which includes mainly the Phishing attacks and many other attacks in the prevailing COVID -19 situation, have reached an alarming rate with the outburst of numerous forms of crime. This paper focuses on various types of cyber crime and targets some of the present day cyber crime attacks based on Phishing, Artificial Intelligence, Cloud technology and Block chain. The principal objective of this work is to identify how Machine Learning can be deployed in detection of diversified fields of cyber crime. The application of various Machine Learning models in the prediction, identification and mitigation of complex threats is also discussed.
This work applies an unsupervised deep feature learning to finding patterns of interharmonics. The main objectives of this work are to provide an additional graphical tool to handle two distinct data inputs: (a) indiv...
详细信息
This work applies an unsupervised deep feature learning to finding patterns of interharmonics. The main objectives of this work are to provide an additional graphical tool to handle two distinct data inputs: (a) individual interharmonics components in time-series; (b) broadband spectrum by employing spectrograms. Both data inputs are analysed employing an autoencoder based on convolutional neural networks followed by clustering. The application of the method results in the most common patterns in time-series or spectrograms. Two study cases are presented by applying the method to measurements from solar installations in Finland and Sweden. The results show the usefulness of the method to recognize interharmonics in a single frequency and broadband spectrum.
Sequential pattern mining has gained popularity in Data Mining and patternrecognition. Most sequential pattern mining algorithms are influenced by noisy variables, parameters tuning, bias-variance dilemma and learnin...
详细信息
Sequential pattern mining has gained popularity in Data Mining and patternrecognition. Most sequential pattern mining algorithms are influenced by noisy variables, parameters tuning, bias-variance dilemma and learning instability. This paper presents a new deep learning model for sequential pattern mining, by using ensemble learning and models selection. Experimental studies on mobile activity recognition showed that our deep learning model, which is named Deep Sequential pattern Mining (abbreviated as DeepSPM), obtained an enhanced generalization in comparison with Long Short Term Memory (LSTM), Bidirectional Associative Memory (BAM) and Hopfield. We provide a comparative performance analysis of patternrecognition. The advantages and the drawbacks of the benchmarking models are critically discussed. (C) 2019 The Authors. Published by Elsevier B.V.
Automatic vending machine is a necessity at this technological era. It's a step to go toward the vendor-less shop management, which supports the commandment of the 4th Industrial Revolution. But image recognition ...
详细信息
Defining methods which map the features present in the input to a binomial class output is the core objective of pattern classification. The Modified Quick Fuzzy Hypersphere Neural Network (MQFHSNN) algorithm proposed...
详细信息
This study presents the development of a wearable device that merges capacitive soft-flexion and surface electromyography (sEMG) sensors for the estimation of shoulder orientation and movement, evaluating five natural...
详细信息
暂无评论