Data Representation Method Research Articles

Deep random walk of unitary invariance for large-scale data representation

Data representation aims at learning an efficient low-dimensional representation, which is always a challenging task in machine learning and computer vision. It can largely improve the performance of specific learning tasks. Unsupervised methods are extensively applied to data representation, which considers the internal connection among data. Most of existing unsupervised models usually use a specific norm to favor certain distributions of the input data, leading to an unsustainable encouraging performance for given learning tasks. In this paper, we propose an efficient data representation method to address large-scale feature representation problems, where the deep random walk of unitary invariance is exploited for learning discriminative features. First, the data representation is formulated as deep random walk problems, where unitarily invariant norms are employed to capture diverse beneficial perspectives hidden in the data. It is embedded into a state transition matrix model, where an arbitrary number of transition steps is available for an accurate affinity evaluation. Second, data representation problems are then transformed as high-order matrix factorization tasks with unitary invariance. Third, a closed-form solution is proved for the formulated data representation problem, which may provide a new perspective for solving high-order matrix factorization problems. Finally, extensive comparative experiments are conducted in publicly available real-world data sets. In addition, experimental results demonstrate that the proposed method achieves better performance than other compared state-of-the-art approaches in terms of data clustering.

Detecting Electricity Theft Cyber-Attacks in AMI Networks Using Deep Vector Embeddings

Despite being equipped with advanced metering infrastructure (AMI), utility companies are subjected to electricity theft cyber-attacks. The existing machine learning-based detectors do not capture well the complex patterns and the temporal correlation present in the time-series profile of energy consumption data. This article proposes a deep recurrent vector embedding model to identify electricity theft cyber-attacks. Vector embedding is a data representation method that we use to express energy consumption profiles as vectors of real numbers. Since the reported electricity readings may be benign or malicious, vector embedding algorithms help in analyzing the relationships and capturing the patterns within the customer's reported readings. Furthermore, our model captures well the time-series nature of the data due to the adoption of gated recurrent units. We implement a sequential grid-search hyperparameter optimization algorithm to further improve the models detection performance. We test our model against two real datasets of benign and malicious readings. Results are 95.8% in detection rate (DR), 2.1% in false alarm (FA), and 93.7% in highest difference (HD). Our model outperforms shallow detectors by 3.5%-9.7% in DR, 3.1%-10% in FA, and 8.7%-21.8% in HD. It also outperforms deep detectors by 1.5%-3.2% in DR, 2%-4.3% in FA, and 5.6%-9.6% in HD.

Classification and Regression of Binary Hydrocarbon Mixtures using Single Metal Oxide Semiconductor Sensor With Application to Natural Gas Detection

To detect natural gas pipeline leaks, ethane in the natural gas can be used to discriminate it from background methane emissions. Our approach classifies samples of methane and simulated natural gas based on time-series data collected from a low-cost microfluidic detector. This detector houses a single metal oxide semiconducting (MOS) gas sensor in a coated microchannel, which provides selectivity based on diffusion times of different gases. A data generation apparatus was designed and built, allowing us to create customized gas mixtures and collect data from our sensing apparatus in an automated fashion. We present a comparison of machine learning models and data representation methods, and demonstrate the feasibility of both discriminating methane from simulated natural gas mixtures, and also obtaining accurate concentration estimates in random binary mixtures of methane and ethane. We achieve a 98.75% classification rate between methane, ethane, and binary mixtures, and a 12.0% mean relative error in regression estimates for arbitrary mixtures. In addition, we discriminate samples of pure methane from simulated natural gas mixtures containing 1% and 3% ethane with 86.7% and 93.3% accuracy, respectively, and regress the concentrations of both the methane and ethane components with a maximum of a 19.3% mean relative error.

Robust sparse coding via self-paced learning for data representation

Sparse coding (SC), due to its thorough theoretical property and outstanding effectiveness, is attracting more and more attention in various data representation and data mining applications. However, the optimization of most existing sparse coding algorithms are non-convex and thus prone to become stuck into bad local minima under the framework of alternative optimization, especially when there are many outliers and noisy data. To enhance the learning robustness, in this study, we will present an unified framework named Self-Paced Sparse Coding (SPSC), which gradually includes data into the learning process of SC from easy ones to complex ones by incorporating self-paced learning methodology. It implements a soft instance selection accordingly rather than a heuristic hard strategy sample selection. We also generalize the self-paced learning schema into different levels of dynamic selection on instances, features and elements respectively. Further, we show an optimization algorithm to solve it and a theoretical explanation to analyze the effectiveness of it. Extensive experimental results on the real-world clean image datasets and images with two kinds of corruptions demonstrate the remarkable robustness of the proposed method for high dimensional data representation on image clustering and reconstruction tasks over the state-of-the-arts.

A computing model for trend analysis in stock data stream classification

<p>For several decades, many statistical and scientific efforts took place for the better analysis or prediction of stock trading. But still it is open to offer new avenues for the scientists to rethink and discover new inferences by adopting latest technological scenarios. In this regard, this paper is trying to apply classification techniques on stock data stream through feature extraction for the trend analysis. The proposed work is involving k-means for clustering samples into two clusters (the stocks in trend as one cluster and another on as stocks not in trend). The trend analysis is done based on density estimation of the stocks with respect to sectors. A well-known data representation method that is histogram is used to represent the sector which is in trend. This work has been implemented and experimented by considering live NSE (india) data using python and its related tools.</p>

Data representation for CNN based internet traffic classification: a comparative study

It has been well established that the Internet of Things will bring an expansion in traffic volume and types. This will bring new challenges in terms of Quality of Service (QoS) and security, requiring innovative traffic management techniques. Traffic classification is a main network function that helps in managing both QoS and security. Different machine learning based methods have been applied for this aim. However, traditional machine learning methods rely on hand crafted features, limiting the model ability to learn. Deep Learning (DL), a branch of machine learning, is characterized by its representation learning ability. In this paper, we analyse two methods of data representation for DL-based classification: a raw packet-based representation and a quasi-raw flow-based representation. Different tests are performed to evaluate the robustness of these data representation methods. The tests include features’ importance, model robustness, and anonymization tests. The results show that raw data representation suffers from traffic anonymization and the fact that many packet fields are data-dependent. On the other hand, the flow-based representation is sensitive to the number of packets used for classification and to traffic obfuscation.

Knowledge-driven drug repurposing using a comprehensive drug knowledge graph.

Due to the huge costs associated with new drug discovery and development, drug repurposing has become an important complement to the traditional de novo approach. With the increasing number of public databases and the rapid development of analytical methodologies, computational approaches have gained great momentum in the field of drug repurposing. In this study, we introduce an approach to knowledge-driven drug repurposing based on a comprehensive drug knowledge graph. We design and develop a drug knowledge graph by systematically integrating multiple drug knowledge bases. We describe path- and embedding-based data representation methods of transforming information in the drug knowledge graph into valuable inputs to allow machine learning models to predict drug repurposing candidates. The evaluation demonstrates that the knowledge-driven approach can produce high predictive results for known diabetes mellitus treatments by only using treatment information on other diseases. In addition, this approach supports exploratory investigation through the review of meta paths that connect drugs with diseases. This knowledge-driven approach is an effective drug repurposing strategy supporting large-scale prediction and the investigation of case studies.

A review on machine learning–based approaches for Internet traffic classification

Traffic classification acquired the interest of the Internet community early on. Different approaches have been proposed to classify Internet traffic to manage both security and Quality of Service (QoS). However, traditional classification approaches consisting of modifying the Transmission Control Protocol/Internet Protocol (TCP/IP) scheme have not been adopted due to their complex management. In addition, port-based methods and deep packet inspection have limitations in dealing with new traffic characteristics (e.g., dynamic port allocation, tunneling, encryption). Conversely, machine learning (ML) solutions effectively classify traffic down to the device type and specific user action. Another research direction aims to anonymize Internet traffic and thwart classification to maintain user privacy. Existing traffic surveys focus on classification and do not consider anonymization. Here, we review the Internet traffic classification and obfuscation techniques, largely considering the ML-based solutions. In addition, this paper presents a comprehensive review of various data representation methods, and the different objectives of Internet traffic classification. Finally, we present the key findings, limitations, and recommendations for future research.

Temporal tree representation for similarity computation between medical patients.

The aim of this study is to compute similarities between patient records in an electronic health record (EHR). This is an important problem because the availability of effective methods for the computation of patient similarity would allow for assistance with and automation of tasks such as patients stratification, medical prognosis and cohort selection, and for unlocking the potential of medical analytics methods for healthcare intelligence. However, health data in EHRs presents many challenges that make the automatic computation of patient similarity difficult; these include: temporal aspects, multivariate, heterogeneous and irregular data, and data sparsity. We propose a new method for EHR data representation called Temporal Tree: a temporal hierarchical representation which, based on temporal co-occurrence, preserves the compound information found at different levels in health data. In addition, this representation is augmented using the doc2vec embedding technique which here is exploited for patient similarity computation. We empirically investigate our proposed method, along with several state-of-the-art benchmarks, on a dataset of real world Intensive Care Unit (ICU) EHRs, for the task of identifying patients with a specific target diagnosis. Our empirical results show that the Temporal Trees representation is significantly better than other traditional and state-of-the-art methods for representing patients and computing their similarities. Temporal trees capture the temporal relationships between medical, hierarchical data: this enables to effectively model the rich information provided within EHRs and thus the identification of similar patients.

Quantum and classical machine learning for the classification of non-small-cell lung cancer patients

The ability to accurately classify disease subtypes is of vital importance, especially in oncology where this capability could have a life saving impact. Here we report a classification between two subtypes of non-small cell lung cancer, namely Adenocarcinoma versus Squamous cell carcinoma. The data consists of approximately 20,000 gene expression values for each of 104 patients. The data was curated from Kuner (Geo lung cancer data set—gse10245, 2009, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE10245 ) and Golumbic (Geo lung cancer data set—gse18842, 2010, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE18842 ). We used a combination of classical and quantum machine learning models to successfully classify these patients. We utilized feature selection methods based on univariate statistics in addition to XGBoost (Guestrin and Chen in Xgboost: a scalable tree boosting system, 2007, https://dl.acm.org/citation.cfm?id=1273596 ). A novel and proprietary data representation method developed by one of the authors called QCrush was also used as it was designed to incorporate a maximal amount of information under the size constraints of the D-Wave quantum annealing computer. The machine learning was performed by a quantum Boltzmann machine. This paper will report our results, the various classical methods, and the quantum machine learning approach we utilized.

Distributed situation awareness: a health-system approach to assessing and designing patient flow management

Patient flow management is a system-wide process but many healthcare providers do not integrate multiple departments into the process to minimise the time between treatments or medical services for maximum patient throughput. This paper presents a case study of applying Distributed Situation Awareness (DSA) to characterise system-wide patient flow management and identify opportunities for improvements in a healthcare system. This case study employed a three-part method of data elicitation, extraction, and representation to investigate DSA. Social, task, and knowledge networks were developed and then combined to characterise patient flow management and identify deficiencies of the command and control centre of a healthcare facility. Social network analysis provided centrality metrics to further characterise patient flow management. The DSA model helped identify design principles and deficiencies in managing patient flow. These findings indicate that DSA is promising for analysing patient flow management from a system-wide perspective. Practitioner summary: This article examines Distribution Situation Awareness (DSA) as an analysis framework to study system-wide patient flow management. The DSA yields social, task, and knowledge networks that can be combined to characterise patient flow and identify deficiencies in the system. DSA appears promising for analysing communication and coordination of complex systems. Abbreviations: CDM: critical decision method; CTaC: carilion transfer and communications center; EAST: event analysis systematic teamwork; ED: emergency department; DES: discrete event simulation; DSA: distributed situation awareness; SA: situation awareness; SNA: social network analysis

Study of Data Representation Methods for TCM Clinical Assistant Diagnosis

The unbalanced distribution of medical resources renders the research on TCM (Traditional Chinese Medicine) clinical assistant diagnosis more important to regions with less medical resources. In recent years, more and more clinical assistant diagnosis methods are deep learning (DL) based. The input data representation of these DL models is one of the most important factors for achieving better results. In this paper, different data representations methods are investigated using a multi-layer perceptron for a multi-class multi-label TCM clinical assistant diagnosis task. From the experimental results, it can be concluded that fast-Text representation is more suitable to this task since TCM clinical records are brief with limited information.

GIS “Spatial Mobility of the Szlachta under Peter the First” in a Prosopographic Study

The article discusses possibilities of geochronological tracking technology for studying the spatial mobility of social groups in Russia in the past. The GIS proposed is necessary to visualize and analyze spatial data in a prosopographic research of about 400 szlachta representatives in Peter’s Epoch. Spatial mobility is understood as the intensity of person's translocation through settlements and his ability to respond to external challenges by moving. The archival materials of the Senate inspection of the szlachta (1721-1723) served the basis for the study and the resource formation. Particular attention is paid to the design of software research tools – the PostgreSQL database and the web GIS based on the latter. It is the first time when geochronological tracking as a geoinformatics method was used to prosopographically study the Russian nobility. The methods of historical source spatial data representation and visualization are implemented in the form of a geodatabase that is publicly available. Two program modules (the GIS among them) grant a wide range of Internet users an access to historical sources text data as well as synchronically visualized data on the szlachta service under Peter the first.  The authors conclude that it is promising to create a special web interface which provides users with flexible text and geodata filtering and analysis. The web project created can be used both for research in the field of social history, historical geography, genealogy and for educational purposes in such courses as “historical computer science” and “digital humanities”.

Monte Carlo kilovoltage X-ray tube simulation: A statistical analysis and compact simulation method

IntroductionMonte Carlo (MC) simulations are a powerful tool for improving image quality in X-ray based imaging modalities. An accurate X-ray source model is essential to MC modeling for CBCT but can be difficult to implement on a GPU while maintaining efficiency and memory limitations. A statistical analysis of the photon distribution from a MC X-ray tube simulation is conducted in hopes of building a compact source model. Materials & methodsMC simulations of an X-ray tube were carried out using BEAMnrc. The resulting photons were sorted into four categories: primary, scatter, off-focal radiation (OFR), and both (scatter and OFR). A statistical analysis of the photon components (energy, position, direction) was completed. A novel method for a compact (memory efficient) representation of the PHSP data was implemented and tested using different statistical based linear transformations (PCA, ZCA, ICA), as well as a geometrical transformation. ResultsThe statistical analysis showed all photon groupings had strong correlations between position and direction, with the largest correlation in the primary data. The novel method was successful in compactly representing the primary (error < 2%) and scatter (error < 6%) photon groupings by reducing the component correlations. Discussion & conclusionStatistical linear transforms provide a method of reducing the memory required to accurately simulate an X-ray source in a GPU MC system. If all photon types are required, the proposed method reduces the memory requirements by 3.8 times. When only primary and scatter data is needed, the memory requirement is reduced from gigabytes to kilobytes.

Smart socks: new effective method of gait monitoring for systems with limited number of plantar sensors

Advanced version of DAid® Pressure Sock System (DPSS) for walking/running gait monitoring is proposed. The system enables to study in details time –related parameters of gait, such as foot strike time, stride/swing time, gradients of plantar loading/unloading, etc. both indoor and outdoor. A new method of data processing, analysis and representation is developed for such systems with limited quantity of sensors and implemented in DPSS software. The method considers the foot loading as a propagation of the plantar pressure wave, where separate sensors record amplitudes of pressure waves as reference points. Data processing includes scaling procedure that is based on application of sample entropy as the measure of sensors activity. Processed data are represented in the form of loading wave plots (WPs). WPs show load propagation across the sensors during the gait/stance. Developed system and method were tested for different types of walk/run and corresponding WPs were constructed and evaluated by sport medicine doctors. Good agreement between WPs – based analysis and diagnoses made by direct observation during walk/run tests confirms system efficacy and validity of results. Proposed method essentially simplifies gait studies and interpretation of the results.

Freud: A software suite for high throughput analysis of particle simulation data

The freud Python package is a library for analyzing simulation data. Written with modern simulation and data analysis workflows in mind, freud provides a Python interface to fast, parallelized C++ routines that run efficiently on laptops, workstations, and supercomputing clusters. The package provides the core tools for finding particle neighbors in periodic systems, and offers a uniform API to a wide variety of methods implemented using these tools. As such, freud users can access standard methods such as the radial distribution function as well as newer, more specialized methods such as the potential of mean force and torque and local crystal environment analysis with equal ease. Rather than providing its own trajectory data structure, freud operates either directly on NumPy arrays or on trajectory data structures provided by other Python packages. This design allows freud to transparently interface with many trajectory file formats by leveraging the file parsing abilities of other trajectory management tools. By remaining agnostic to its data source, freud is suitable for analyzing any particle simulation, regardless of the original data representation or simulation method. When used for on-the-fly analysis in conjunction with scriptable simulation software such as HOOMD-blue, freud enables smart simulations that adapt to the current state of the system, allowing users to study phenomena such as nucleation and growth. Program summaryProgram Title:freudProgram Files doi:http://dx.doi.org/10.17632/v7wmv9xcct.1Licensing provisions: BSD 3-ClauseProgramming language: Python, C++Nature of problem: Simulations of coarse-grained, nano-scale, and colloidal particle systems typically require analyses specialized to a particular system. Certain more standardized techniques – including correlation functions, order parameters, and clustering – are computationally intensive tasks that must be carefully implemented to scale to the larger systems common in modern simulations.Solution method:freud performs a wide variety of particle system analyses, offering a Python API that interfaces with many other tools in computational molecular sciences via NumPy array inputs and outputs. The algorithms in freud leverage parallelized C++ to scale to large systems and enable real-time analysis. The library’s broad set of features encode few assumptions compared to other analysis packages, enabling analysis of a broader class of data ranging from biomolecular simulations to colloidal experiments.Additional comments including restrictions and unusual features:1. freud provides very fast parallel implementations of standard analysis methods like RDFs and correlation functions.2. freud includes the reference implementation for the potential of mean force and torque (PMFT).3. freud provides various novel methods for characterizing particle environments, including the calculation of descriptors useful for machine learning. The source code is hosted on GitHub (https://github.com/glotzerlab/freud), and documentation is available online (https://freud.readthedocs.io/). The package may be installed via pip install freud-analysis or conda install -c conda-forge freud.

Investigating Phase Space Reconstruction of ECG for Prediction of Malignant Ventricular Arrhythmia

Prediction of malignant ventricular arrhythmia is imperative to enable early diagnosisand prevent sudden cardiac death. There are a wide variety of methods employed in previous works with box counting of phase space reconstruction diagrams achieving the highest performance. However, there is no follow-up work investigating or improving this method. This work is performed to objectively assess the feasibility of box counting technique for prediction of malignant ventricular arrhythmia and to validate the prediction technique across larger data sets, including the widely acceptable MIT-BIH database. Box counting using different windowing methods, data representation methods of phase portrait and testing database are investigated for performance and versatility. By using windows of RR segments and linking phase portraits, this modified box counting method has higher resistance to signal noises, which are common in the electrocardiogram. It is verified using four Physionet databases (CUDB, SDDB, PTBDB and NSRDB) which contain either arrhythmic or control records. High prediction accuracy of 94.12%, sensitivity of 88.46% and specificity of 100% are achieved by the coefficient of variation derived from this method. This technique is proven useful in predicting malignant ventricular arrhythmia and its implementation is envisaged to enable early detection and diagnosis.

A graph-based taxonomy of citation recommendation models

Recommender systems have been used since the beginning of the Web to assist users with personalized suggestions related to past preferences for items or products including books, movies, images, research papers and web pages. The availability of millions research articles on various digital libraries makes it difficult for a researcher to find relevant articles to his/er research. During the last years, a lot of research have been conducted through models and algorithms that personalize papers recommendations. With this survey, we explore the state-of-the-art citation recommendation models which we categorize using the following seven criteria: platform used, data factors/features, data representation methods, methodologies and models, recommendation types, problems addressed, and personalization. In addition, we present a novel k-partite graph-based taxonomy that examines the relationships among surveyed algorithms and corresponding k-partite graphs used. Moreover, we present (a) domain’s popular issues, (b) adopted metrics, and (c) commonly used datasets. Finally, we provide some research trends and future directions.

Comparative study of landslide susceptibility mapping with different recurrent neural networks

This paper aims to use recurrent neural networks (RNNs) to perform landslide susceptibility mapping in Yongxin County, China. The two main contributions of this study are summarized as follows. First, the regular RNN is compared to its three variants in the case study of landslide susceptibility mapping for the first time, including long short term memory, gated recurrent unit and simple recurrent unit. Second, a sequential data representation method is proposed to fully explore the predicting potential of RNNs. The study area consists of 364 historical landslide locations that were divided into two parts: 255 (70%) for training and 109 (30%) for validation, and 16 landslide influencing factors were considered for spatial prediction. To validate the effectiveness of these RNN-related methods, several objective measures of accuracy, recall, F-measure, Matthews correlation coefficient and the receiver operating characteristic were used for evaluation. Experimental results demonstrate that very high and high susceptible areas are concentrated in the northwest and south of Yongxin County, while landslides in the central area are less prone to occur. Based on quantitative results, all the RNN-related methods achieved area under the curve values above 0.83 and produced accurate prediction results with the optimized parameters. Therefore, the RNN framework can be used as a useful tool for the landslide susceptibility mapping task to mitigate and manage landslides.

An Active Data Representation of Videos for Automatic Scoring of Oral Presentation Delivery Skills and Feedback Generation

Public speaking is an important skill, the acquisition of which requires dedicated and time consuming training. In recent years, researchers have started to investigate automatic methods to support public speaking skills training. These methods include assessment of the trainee's oral presentation delivery skills which may be accomplished through automatic understanding and processing of social and behavioural cues displayed by the presenter. In this study, we propose an automatic scoring system for presentation delivery skills using a novel active data representation method to automatically rate segments of a full video presentation. While most approaches have employed a two step strategy consisting of detecting multiple events followed by classification, which involve the annotation of data for building the different event detectors and generating a data representation based on their output for classification, our method does not require event detectors. The proposed data representation is generated unsupervised using low-level audiovisual descriptors and self-organizing mapping, and used for video classification. This representation is also used to analyse video segments within a full video presentation in terms of several characteristics of the presenter's performance. The audio representation provides the best prediction results for self-confidence and enthusiasm, posture and body language, structure and connection of ideas, and overall presentation delivery. The video data representation provides the best results for presentation of relevant information with good pronunciation, usage of language according to audience , and maintenance of adequate voice volume for the audience. The fusion of audio and video data provides the best results for eye contact. Applications of the method to provision of feedback to teachers and trainees are discussed.