Fast and Accurate Deep Learning Framework for Secure Fault Diagnosis in the Industrial Internet of Things

This article introduced a new deep learning framework for fault diagnosis in electrical power systems. The framework integrates the convolution neural network and different regression models to visually identify which faults have occurred in electric power systems. The approach includes three main steps: 1) data preparation;2) object detection;and 3) hyperparameter optimization. Inspired by deep learning and evolutionary computation (EC) techniques, different strategies have been proposed in each step of the process. In addition, we propose a new hyperparameters optimization model based on EC that can be used to tune parameters of our deep learning framework. In the validation of the framework's usefulness, experimental evaluation is executed using the well known and challenging VOC 2012, the COCO data sets, and the large NESTA 162-bus system. The results show that our proposed approach significantly outperforms most of the existing solutions in terms of runtime and accuracy.

Fuzzy Contrast Set Based Deep Attention Network for Lexical Analysis and Mental Health Treatment

Internet-delivered psychological treatments (IDPT) consider mental problems based on Internet interaction. With such increased interaction because of the COVID-19 pandemic, more online tools have been widely used to provide evidence-based mental health services. This increase helps cover more population by using fewer resources for mental health treatments. Adaptivity and customization for the remedy routine can help solve mental health issues quickly. In this research, we propose a fuzzy contrast-based model that uses an attention network for positional weighted words and classifies mental patient authored text into distinct symptoms. After that, the trained embedding is used to label mental data. Then the attention network expands its lexicons to adapt to the usage of transfer learning techniques. The proposed model uses similarity and contrast sets to classify the weighted attention words. The fuzzy model then uses the sets to classify the mental health data into distinct classes. Our method is compared with non-embedding and traditional techniques to demonstrate the proposed model. From the experiments, the feature vector can achieve a high ROC curve of 0.82 with problems associated with nine symptoms.

An Intelligent Collaborative Image-Sensing System for Disease Detection

With the growth of smart medical devices and applications in smart hospitals, home care facilities, nursing, and the Internet of Medical Things (IoMT) are becoming more ubiquitous. It uses smart medical devices and cloud computing services, and basic Internet of Things (IoT) technology, to detect key body indicators, monitor health situations, and generate multivariate data to provide just-in-time healthcare services. In this article, we present a novel collaborative disease detection system based on IoMT amalgamated with captured image data. The system can be based on intelligent agents, where every agent explores the interaction between different medical data obtained by smart sensor devices using reinforcement learning as well as targets to detect diseases. The agents then collaborate to make a reliable conclusion about the detected diseases. Intensive experiments were conducted using medical data. The results show the importance of using intelligent agents for disease detection in healthcare decision-making. Moreover, collaboration increases the detection rate, with numerical results showing the superiority of the proposed framework compared with baseline solutions for disease detection. © 2001-2012 IEEE.

A Deep Learning based Feature Entity Relationship Extraction Method for Telemedicine Sensing Big Data

To solve the problem of inaccurate entity extraction caused by low application efficiency and big data noise in telemedicine sensing data, a deep learning-based method for entity relationship extraction in telemedicine big data is proposed. By analyzing the distribution structure of the medical sensing big data, the fuzzy function of the distribution shape is calculated and the seed relationship set is transformed by the inverse Shearlet transform. Combined with the deep learning technology, the GMM-GAN data enhancement model is built, the interactive medical sensing big data features are obtained, the association rules are matched one by one, the noiseless medical sensing data are extracted in time sequence, the feature items with the highest similarity are obtained and used as the constraint to complete the feature entity relationship extraction of the medical sensing data. The experimental results show that the extracted similarity of entity relations is more than 70%, which can handle overly long and complex sentences in telemedicine information text;the extraction time is the shortest and the volatility is low.

Toward a Cognitive-Inspired Hashtag Recommendation for Twitter Data Analysis

This research investigates hashtag suggestions in a heterogeneous and huge social network, as well as a cognitive-based deep learning solution based on distributed knowledge graphs. Community detection is first performed to find the connected communities in a vast and heterogeneous social network. The knowledge graph is subsequently generated for each discovered community, with an emphasis on expressing the semantic relationships among the Twitter platform's user communities. Each community is trained with the embedded deep learning model. To recommend hashtags for the new user in the social network, the correlation between the tweets of such user and the knowledge graph of each community is explored to set the relevant communities of such user. The models of the relevant communities are used to infer the hashtags of the tweets of such users. We conducted extensive testing to demonstrate the usefulness of our methods on a variety of tweet collections. Experimental results show that the proposed approach is more efficient than the baseline approaches in terms of both runtime and accuracy.

COVID-19 Genome Analysis Using Alignment-Free Methods

Examining the genome sequences of the novel coronavirus (COVID-19) strains is critical to properly understand this disease and its functionalities. In bioinformatics, alignment-free (AF) sequence analysis methods offer a natural framework to investigate and understand the patterns and inherent properties of biological sequences. Thus, AF methods are used in this paper for the analysis and comparison of COVID-19 genome sequences. First, frequent patterns of nucleotide base(s) in COVID-19 genome sequences are extracted. Second, the similarity/dissimilarity between COVID-19 genome sequences are measured with different AF methods. This allows to compare sequences and evaluate the performance of various distance measures employed in AF methods. Lastly, the phylogeny for the COVID-19 genome sequences are constructed with various AF methods as well as the consensus tree that shows the level of support (agreement) among phylogenetic trees built by various AF methods. Obtained results show that AF methods can be used efficiently for the analysis of COVID-19 genome sequences. © 2021, Springer Nature Switzerland AG.