DeepQFM: a deep learning based query facets mining method

Detecting the technology's evolutionary pathway using HiDS-trait-driven tech mining strategy

Recently, deep learning has developed rapidly, while it has also been quite successfully applied in the field of hyperspectral classification. Generally, training the parameters of a deep neural network to the best is the core step of a deep learning-based method, which usually requires a large number of labeled samples. However, in remote sensing analysis tasks, we only have limited labeled data because of the high cost of their collection. Therefore, in this paper, we propose a deep metric learning with online hard mining (DMLOHM) method for hyperspectral classification, which can maximize the inter-class distance and minimize the intra-class distance, utilizing a convolutional neural network (CNN) as an embedded network. First of all, we utilized the triplet network to learn better representations of raw data so that raw data were capable of having their dimensionality reduced. Afterward, an online hard mining method was used to mine the most valuable information from the limited hyperspectral data. To verify the performance of the proposed DMLOHM, we utilized three well-known hyperspectral datasets: Salinas Scene, Pavia University, and HyRANK for verification. Compared with CNN and DMLTN, the experimental results showed that the proposed method improved the classification accuracy from 0.13% to 4.03% with 85 labeled samples per class.

Due to the large amount of data information in the Gas insulated switchgear (GIS) partial discharge defect sample database, if the analysis and diagnosis are directly carried out, there will be problems of slow data mining speed, poor efficiency and low accuracy. Therefore, further research is needed to apply to different types of insulation partial discharge data abnormalities of GIS equipment. The typical feature mining and fault diagnosis method of the state. First, Use the Association analysis rules and the neural network to make a data cleaning strategy to clean and normalize the sample data. Then use the self-learning algorithm evolution mechanism to train the deep neural network, update the model parameters, and continuously modify the abnormal state data of the sample library, and build the GIS equipment insulation defect partial discharge data sample knowledge base. Finally, a deep belief network partial discharge signal diagnosis neural network learning model and optimization technology based on deep learning are constructed to improve the adaptability and generalization of different types of partial discharge signal map diagnosis of GIS equipment, as well as the intelligent diagnosis speed and abnormal state information accuracy.

BackgroundData measuring airborne pollutants, public health and environmental factors are increasingly being stored and merged. These big datasets offer great potential, but also challenge traditional epidemiological methods. This has motivated the exploration of alternative methods to make predictions, find patterns and extract information. To this end, data mining and machine learning algorithms are increasingly being applied to air pollution epidemiology.MethodsWe conducted a systematic literature review on the application of data mining and machine learning methods in air pollution epidemiology. We carried out our search process in PubMed, the MEDLINE database and Google Scholar. Research articles applying data mining and machine learning methods to air pollution epidemiology were queried and reviewed.ResultsOur search queries resulted in 400 research articles. Our fine-grained analysis employed our inclusion/exclusion criteria to reduce the results to 47 articles, which we separate into three primary areas of interest: 1) source apportionment; 2) forecasting/prediction of air pollution/quality or exposure; and 3) generating hypotheses. Early applications had a preference for artificial neural networks. In more recent work, decision trees, support vector machines, k-means clustering and the APRIORI algorithm have been widely applied. Our survey shows that the majority of the research has been conducted in Europe, China and the USA, and that data mining is becoming an increasingly common tool in environmental health. For potential new directions, we have identified that deep learning and geo-spacial pattern mining are two burgeoning areas of data mining that have good potential for future applications in air pollution epidemiology.ConclusionsWe carried out a systematic review identifying the current trends, challenges and new directions to explore in the application of data mining methods to air pollution epidemiology. This work shows that data mining is increasingly being applied in air pollution epidemiology.The potential to support air pollution epidemiology continues to grow with advancements in data mining related to temporal and geo-spacial mining, and deep learning. This is further supported by new sensors and storage mediums that enable larger, better quality data. This suggests that many more fruitful applications can be expected in the future.

DNA/RNA motif mining is the foundation of gene function research. The DNA/RNA motif mining plays an extremely important role in identifying the DNA- or RNA-protein binding site, which helps to understand the mechanism of gene regulation and management. For the past few decades, researchers have been working on designing new efficient and accurate algorithms for mining motif. These algorithms can be roughly divided into two categories: the enumeration approach and the probabilistic method. In recent years, machine learning methods had made great progress, especially the algorithm represented by deep learning had achieved good performance. Existing deep learning methods in motif mining can be roughly divided into three types of models: convolutional neural network (CNN) based models, recurrent neural network (RNN) based models, and hybrid CNN–RNN based models. We introduce the application of deep learning in the field of motif mining in terms of data preprocessing, features of existing deep learning architectures and comparing the differences between the basic deep learning models. Through the analysis and comparison of existing deep learning methods, we found that the more complex models tend to perform better than simple ones when data are sufficient, and the current methods are relatively simple compared with other fields such as computer vision, language processing (NLP), computer games, etc. Therefore, it is necessary to conduct a summary in motif mining by deep learning, which can help researchers understand this field.

The current conventional sentiment lexicon mining methods mainly achieve lexical extraction by mining the sentiment features of words, which leads to poor mining effect due to the lack of sentiment lexicon construction. In this regard, the intelligent mining method of sentiment lexicon of English translation text based on deep learning is proposed. The intelligent mining algorithm is optimized by extracting lexical templates with generality, constructing sentiment lexicon, and building convolutional network to extract lexical features. The experimental results showed that this mining method had an accuracy of over 85% in mining emotional vocabulary in translated texts, which was higher than other algorithms. Meanwhile, the mining time required for this algorithm was not yet 4.5 min. The above results indicated that the proposed method had higher accuracy and faster speed in mining emotional vocabulary in English translation texts.

Currently, most IT organizations are inclined towards a cloud computing environment because of its distributed and scalable nature. However, its flexible and open architecture is receiving lots of attention from potential intruders for cyber threats. Here, Intrusion Detection System (IDS) plays a significant role in monitoring malicious activities in cloud-based systems. The state of the art of this paper is to systematically review the existing methods for detecting intrusions based upon various techniques, such as data mining, machine learning, and deep learning methods. Recently, deep learning techniques have gained momentum in the intrusion detection domain, and several IDS approaches are provided in the literature using various deep learning techniques to deal with privacy concerns and security threats. For this purpose, the article focuses on the deep IDS approaches and investigates how deep learning networks are employed by different approaches in various steps of the intrusion detection process to achieve better results. Then, it provided a comparison of the deep learning approaches and the shallow machine learning methods. Also, it describes datasets that are most used in IDS.

In data mining and machine learning methods, it is traditionally assumed that training data, test data, and the data that will be processed in the future, should have the same feature space distribution. This is a condition that will not happen in the real world. In order to overcome this challenge, domain adaptation-based methods are used. One of the existing challenges in domain adaptation-based methods is to select the most efficient features so that they can also show the most efficiency in the destination database. In this paper, a new feature selection method based on deep reinforcement learning is proposed. In the proposed method, in order to select the best and most appropriate features, the essential policies in deep reinforcement learning are defined, and then the selection features are applied for training random forest, k-nearest neighborhood and support vector machine classifiers. The trained classifiers with the considered features are evaluated on the target database. The results are evaluated with the criteria of accuracy, sensitivity, positive and negative predictive rates in the classifiers. The achieved results show the superiority of the proposed method of feature selection when used in domain adaptation. By implementing the RF classifier on the VisDA-2018 database and the Syn2Real database, the classification accuracy in the feature selection of the proposed deep learning reinforcement has increased compared to the two-feature selection of Laplace monitoring and feature selection states. The classification sensitivity with the help of SVM classifier on the Syn2Real databases had the highest values in the feature selection state of the proposed deep learning reinforcement. The obtained number 100 is a positive predictive rate in the Syn2Real database with the help of SVM classifier and in the case of selecting the proposed feature, it indicates its superiority. The negative predictive rate in the Syn2Real database in the state of feature selection of the proposed deep reinforcement learning was 100%, which showed its superiority in comparison with 90.1% in the state of selecting the Laplace monitoring feature. Gmean in KNN classifier on the Syn2Real database has improved in the feature selection state of the proposed deep learning reinforcement in comparison to without feature selection state.

Educational Data Mining (EDM) is a research field that focuses on the application of data mining, machine learning, and statistical methods to detect patterns in large collections of educational data. Different machine learning techniques have been applied in this field over the years, but it has been recently that Deep Learning has gained increasing attention in the educational domain. Deep Learning is a machine learning method based on neural network architectures with multiple layers of processing units, which has been successfully applied to a broad set of problems in the areas of image recognition and natural language processing. This paper surveys the research carried out in Deep Learning techniques applied to EDM, from its origins to the present day. The main goals of this study are to identify the EDM tasks that have benefited from Deep Learning and those that are pending to be explored, to describe the main datasets used, to provide an overview of the key concepts, main architectures, and configurations of Deep Learning and its applications to EDM, and to discuss current state‐of‐the‐art and future directions on this area of research.

In view of the poor effect of most fault diagnosis methods on the intelligent recognition of equipment images, a fault diagnosis method of distribution equipment based on the hybrid model of robot and deep learning is proposed to reduce the dependence on manpower and realize efficient intelligent diagnosis. Firstly, the robot is used to collect the on-site state images of distribution equipment to build the image information database of distribution equipment. At the same time, the robot background is used as the comprehensive database data analysis platform to optimize the sample quality of the database. Then, the massive infrared images are segmented based on chroma saturation brightness space to distinguish the defective equipment images, and the defective equipment areas are extracted from the images by OTSU method. Finally, the residual network is used to improve the region-based fully convolutional networks (R-FCN) algorithm, and the improved R-FCN algorithm trained by the online hard example mining method is used for fault feature learning. The fault type, grade, and location of distribution equipment are obtained through fault criterion analysis. The experimental analysis of the proposed method based on PyTorch platform shows that the fault diagnosis time and accuracy are about 5.5 s and 92.06%, respectively, which are better than other comparison methods and provide a certain theoretical basis for the automatic diagnosis of power grid equipment.

Simple SummaryResearch on proteins and their interactions with other proteins yields many new findings that help explain how diseases emerge. However, manual curation of scientific literature delays new discoveries in the field. Artificial intelligence and deep learning techniques have played a significant part in information extraction from textual forms. In this study, we used text mining and artificial intelligence techniques to address the issue of extracting protein–protein interaction networks from the vast amount of scientific research literature. We have created an automated system consisting of three models using deep learning and natural language processing methods. The accuracy of our first model, which employs recurrent neural networks using sentiment analysis, was 95%. Additionally, the accuracy of our second model, which employs the named entity recognition technique in NLP, was effective and achieved an accuracy of 98%. In comparison to the protein interaction network, we discovered by manual curation of more than 30 articles on Autism Spectrum Disorder, that the automated system testing on 6027 abstracts was successful in developing the network of interactions and provided an improved view. Discovering these networks will greatly help physicians and scientists understand how these molecules interact for physiological, pharmacological, and pathological insight.Text mining methods are being developed to assimilate the volume of biomedical textual materials that are continually expanding. Understanding protein–protein interaction (PPI) deficits would assist in explaining the genesis of diseases. In this study, we designed an automated system to extract PPIs from the biomedical literature that uses a deep learning sentence classification model, a pretrained word embedding, and a BiLSTM recurrent neural network with additional layers, a conditional random field (CRF) named entity recognition (NER) model, and shortest-dependency path (SDP) model using the SpaCy library in Python. The automated system ensures that it targets sentences that contain PPIs and not just these proteins mentioned in the framework of disease discovery or other context. Our first model achieved 13% greater precision on the Aimed/BioInfr benchmark corpus than the previous state-of-the-art BiLSTM neural network models. The NER model presented in this study achieved 98% precision on the Aimed/BioInfr corpus over previous models. In order to facilitate the production of an accurate representation of the PPI network, the processes were developed to systematically map the protein interactions in the texts. Overall, evaluating our system through the use of 6027 abstracts pertaining to seven proteins associated with Autism Spectrum Disorder completed the manually curated PPI network for these proteins. When it comes to complicated diseases, these networks would assist in understanding how PPI deficits contribute to disease development while also emphasizing the influence of interactions on protein function and biological processes.

Good representations of data eliminate irrelevant variability of the input data, while preserving the information that is useful for the ultimate task. Among the various ways for learning representation is using deep learning methods. Deep feature hierarchies are formed by stacking unsupervised modules on top of each other, forming multiple nonlinear transformations to produce better representations. In this talk, we will first show how deep learning is used for bioactivity prediction of chemical compounds. Molecules are represented as several convolutional neural networks to predict their bioactivity. In addition, a new concept of merging multiple convolutional neural networks and an automatic learning features representation for the chemical compounds was proposed using the values within neurons of the last layer of the CNN architecture. We will also show how the concepts of deep learning is adapted into a deep belief network (DBN) to enhance the molecular similarity searching. The DBN achieves feature abstraction by reconstruction weight for each feature and minimizing the reconstruction error over the whole feature set. The DBN is later enhanced using data fusion to obtain a lower detection error probability and a higher reliability by using data from multiple distributed descriptors. Secondly, we will show how we used deep learning for stock market prediction. Here, we developed a Deep Long Short Term Memory Network model that is able to forecast the crude palm oil price movement with combined factors such as other commodities prices, weather and news sentiments and price movement of crude palm oil. We also show how we combined stock markets price and financial news and deployed the Long Short Term Memory (LSTM), Recurrent Neural Network (RNN), and Word 2 Vector (Word2Vec) to project the stock prices for the following seven days. Finally, we will show how we exploited deep learning method for the opinion mining and later used it to extract the product's aspects from the user textual review for recommendation systems. Specifically, we employ a multichannel convolutional neural network (MCNN) for two different input layers, namely, word embedding layer and Part-of-speech (POS) tag embedding layer. We show effectiveness of the proposed model in terms of both aspect extraction and rating prediction performance.

Artificial Intelligence and Cellular Segmentation in Tissue Microscopy Images

With developments in human societies and the information and communication technology, the Internet of Things (IoT) has penetrated various aspects of daily life and different industries. The newly emerging blockchain technology has become a viable solution to the IoT security due to its inherent characteristics such as distribution, security, immutability, and traceability. However, integrating the IoT with the blockchain technology faces certain challenges such as latency, throughput, device power limitation, and scalability. Recent studies have focused on the role of artificial intelligence methods in improving the IoT performance in a blockchain. According to their results, there are only a few effects on the improvement of IoT-based performance with limited power. This study proposes a conceptual model to improve the blockchain throughput in IoT-based devices with limited power through deep reinforcement learning. This model benefits from a recommender agent based on deep reinforcement learning in the mobile edge computing layer to improve the throughput and select the right mining method.

The rapid development of software has brought unprecedented severe challenges to software security vulnerabilities. Traditional vulnerability mining methods are difficult to apply to large-scale software systems due to drawbacks such as manual inspection, low efficiency, high false positives and high false negatives. Recent research works have attempted to apply deep learning models to vulnerability mining, and have made a good progress in vulnerability mining filed. In this paper, we analyze the deep learning model framework applied to vulnerability mining and summarize its overall workflow and technology. Then, we give a detailed analysis on five feature extraction methods for vulnerability mining, including sequence characterization-based method, abstract syntax tree-based method, graph-based method, text-based method and mixed characterization-based method. In addition, we summarize their advantages and disadvantages from the angles of single and mixed feature extraction method. Finally, we point out the future research trends and prospects.