Data Dimensionality Reduction Techniques Research Articles

An optimization framework for planning wayside and on-board hybrid storage systems for tramway applications

Electric mobility is gaining importance on pursuing decarbonisation of transport sector. In this context, electric tramways are recognised as a clean urban collective conveyance. Strong evidences manifest the advantages of storage systems in tramway applications. Despite its apparent benefits, the planning of such kind of systems for tramway applications has received few attention in literature. This work aims at filling this gap. To this end, a novel optimization framework for planning hybrid storage systems (batteries + super-capacitors) for tramway applications on either wayside or on-board configurations is developed, which incorporates an energy management tool to effectively coordinate the different storage technologies. Specific issues related with high resolution tramway demand measurements are addressed by using data dimensionality reduction techniques and temporal representation by representative trips. The new proposal is validated for a real Cuenca - Ecuador tramway for which extensive simulations are carried out with different storage system layouts (super-capacitors, super-capacitors + batteries), and configurations (wayside and on-board) including different electrochemical batteries technologies (lead-acid, nickel cadmium and lithium ion). The results obtained show the benefits of storage systems in tramway facilities (in fact, the total cost of the project can be reduced by 71%), highlighting the importance of fast-acting technologies such as super-capacitors to handle high peak demand consumptions, while batteries are primarily dedicated to providing backup long-term storage capacity. Wayside and on-board configurations are also compared, revealing that total weight of the storage system suppose the main barrier for large capacity installation in on-board layouts.

FEATURE EXTRACTION OF DESIS AND PRISMA HYPERSPECTRAL REMOTE SENSING DATASETS FOR GEOLOGICAL APPLICATIONS

Abstract. With the recent launch of advanced hyperspectral satellites with global coverage, including DESIS and PRISMA, a massive volume of high spectral resolution data is available. This work is focused on the spectral analysis and implementation of feature extraction or data dimensionality reduction techniques on both newly available datasets for geological interpretation. Three of the best feature extraction algorithms, Principal Component Analysis (PCA), Minimum Noise Fraction (MNF), and Independent Component Analysis (ICA), were tested for lithological mapping for the Rajasthan state of India. The present work demonstrates the advantage of the feature extraction algorithm in geological mapping and interpretability as it shows the excellent performance for these datasets. The narrowband ratios for the clay minerals, dolomite, kaolinite, amphiboles, and Al-OH are generated using the PCA and MNF components. All of these band ratios were compared with the Lithological Map available. It is concluded that PCA is the first choice for feature-based lithological classification of hyperspectral remote sensing data. ICA is giving some impressive results which can be explored further. DESIS and PRISMA have a 30 km swath, finer spectral resolution, and high signal-to-noise ratio, which shows much potential in lithological mapping over the parts of northern India. It is suggested to use advanced feature extraction algorithms with recently launched hyperspectral data for accurate and updated mineral mapping over a sizeable geological importance area.

Big data analytics in Industry 4.0 ecosystems

Big data analytics in Industry 4.0 ecosystems

A visual analytic approach for the identification of ICU patient subpopulations using ICD diagnostic codes.

A large number of clinical concepts are categorized under standardized formats that ease the manipulation, understanding, analysis, and exchange of information. One of the most extended codifications is the International Classification of Diseases (ICD) used for characterizing diagnoses and clinical procedures. With formatted ICD concepts, a patient profile can be described through a set of standardized and sorted attributes according to the relevance or chronology of events. This structured data is fundamental to quantify the similarity between patients and detect relevant clinical characteristics. Data visualization tools allow the representation and comprehension of data patterns, usually of a high dimensional nature, where only a partial picture can be projected. In this paper, we provide a visual analytics approach for the identification of homogeneous patient cohorts by combining custom distance metrics with a flexible dimensionality reduction technique. First we define a new metric to measure the similarity between diagnosis profiles through the concordance and relevance of events. Second we describe a variation of the Simplified Topological ion of Data (STAD) dimensionality reduction technique to enhance the projection of signals preserving the global structure of data. The MIMIC-III clinical database is used for implementing the analysis into an interactive dashboard, providing a highly expressive environment for the exploration and comparison of patients groups with at least one identical diagnostic ICD code. The combination of the distance metric and STAD not only allows the identification of patterns but also provides a new layer of information to establish additional relationships between patient cohorts. The method and tool presented here add a valuable new approach for exploring heterogeneous patient populations. In addition, the distance metric described can be applied in other domains that employ ordered lists of categorical data.

A systematic review on machine learning for fall detection system

AbstractFall is a major threat to the health and life of the elders. A Fall Detection System (FDS) assist the elders by identifying the fall and save their life. Machine Learning‐ (ML) based FDS has turned into a major research area due to its capability to assist the elders automatically. The efficiency of a FDS depends on its strength to identify the fall from nonfall accurately. The initial fall detection scheme depends on the threshold‐based classification to classify the fall from the Activity of Daily Living (ADL) but this classification method has failed to reduce the false alarm rate, which raises a question on the efficiency of the technique. This review work identifies the problems in threshold‐based classification from existing works and finds the need for an efficient ML‐based classification technique to accurately identify the fall. Then, presents a comprehensive literature review on various ML‐based classification in fall detection. Moreover, the scrutiny investigates the shortcomings associated with the ML‐based techniques for future research. This study finds that present ML‐based FDS has not addressed problems like data preprocessing and data dimensionality reduction techniques even though ML‐based techniques are far superior to threshold‐based techniques. The study concludes that Self‐Adaptive‐based FDS, as well as the complexity reduction of ML‐based models, should be concentrated in future research.

An Alternating Rank-k Nonnegative Least Squares Framework (ARkNLS) for Nonnegative Matrix Factorization

Nonnegative matrix factorization (NMF) is a prominent technique for data dimensionality reduction that has been widely used for text mining, computer vision, pattern discovery, and bioinformatics. In this paper, a framework called ARkNLS (Alternating Rank-k Nonnegativity constrained Least Squares) is proposed for computing NMF. First, a recursive formula for the solution of the rank-k nonnegativity-constrained least squares (NLS) is established. This recursive formula can be used to derive the closed-form solution for the Rank-k NLS problem for any integer k $\ge$ 1. As a result, each subproblem for an alternating rank-k nonnegative least squares framework can be obtained based on this closed form solution. Assuming that all matrices involved in rank-k NLS in the context of NMF computation are of full rank, two of the currently best NMF algorithms HALS (hierarchical alternating least squares) and ANLS-BPP (Alternating NLS based on Block Principal Pivoting) can be considered as special cases of ARkNLS with k = 1 and k = r for rank r NMF, respectively. This paper is then focused on the framework with k = 3, which leads to a new algorithm for NMF via the closed-form solution of the rank-3 NLS problem. Furthermore, a new strategy that efficiently overcomes the potential singularity problem in rank-3 NLS within the context of NMF computation is also presented. Extensive numerical comparisons using real and synthetic data sets demonstrate that the proposed algorithm provides state-of-the-art performance in terms of computational accuracy and cpu time.

Data dimensionality reduction techniques for Industry 4.0: Research results, challenges, and future research directions

SummaryFrom the last few years, we have witnessed the fourth generation industrial revolution (Industry 4.0), impact of which will be seen in the years to come in various disciplines such as healthcare, transportation, IoT, smart grid, autonomous vehicles, and image processing. These applications in Industry 4.0 may have data in the form of images, speech signals, videos having high dimensions containing multiple dimensions to represent data along different axis. So, the complexity of data processing increases with an increase in the dimensions of the dataset. Complexity can be viewed in terms of detecting and exploiting the relationships among different features of the dataset. These complexities among different attributes can be reduced with the help of dimensionality reduction techniques. These techniques reduce the dimensions from the original input dataset to a lower dimensional dataset. Dimensionality reduction methods are broadly categorized into two types asfeature extraction and feature selection. In feature selection method, out of the original set, a subset of features are identified to get a smaller subset which can be used to build the model whereas, the feature extraction method reduces the dataset of high dimensions to a lower dimension space, that is, a space with a less number of features having different values in comparison to the original dataset. Keeping focus on these points, in this article, we have compared and analyzed different data dimensionality reduction techniques which reduce the dimensions of a large and complex dataset during data processing. In addition, we have discussed various data dimensionality reduction techniques and compared these techniques with respect to various parameters. The comparison among various techniques provides insights to the readers about the applicability of a specific technique to the stand‐alone or a group of applications.

Towards a Machine-Learning-Assisted Dielectric Sensing Platform for Point-of-Care Wound Monitoring

In this letter, we present a machine-learning-based solution to classify wounds and normal skin based on a dielectric spectroscopy approach. Using a commercial network analyzer, we have measured the dielectric constant of normal skin and different types of wounds from multiple living mice across a frequency range from 10 MHz to 20 GHz. The acquired data across a wide frequency range is processed by a Data Dimensionality Reduction technique to extract the optimum frequency for wound dielectric spectroscopy. The results of our analysis reveal that different types of wounds can be distinguished by acquiring the dielectric constants in a frequency range from 1 to 2 GHz. This finding relaxes the large bandwidth requirements of dielectric spectroscopy sensors. By adopting supervised learning classification tools, we have demonstrated that various tissue types across different samples can be classified with an accuracy of near 100%.

Efficient regularized spectral data embedding

Data embedding (DE) or dimensionality reduction techniques are particularly well suited to embedding high-dimensional data into a space that in most cases will have just two dimensions. Low-dimensional space, in which data samples (data points) can more easily be visualized, is also often used for learning methods such as clustering. Sometimes, however, DE will identify dimensions that contribute little in terms of the clustering structures that they reveal. In this paper we look at regularized data embedding by clustering, and we propose a simultaneous learning approach for DE and clustering that reinforces the relationships between these two tasks. Our approach is based on a matrix decomposition technique for learning a spectral DE, a cluster membership matrix, and a rotation matrix that closely maps out the continuous spectral embedding, in order to obtain a good clustering solution. We compare our approach with some traditional clustering methods and perform numerical experiments on a collection of benchmark datasets to demonstrate its potential.

Structured optimal graph based sparse feature extraction for semi-supervised learning

Graph-based feature extraction is an efficient technique for data dimensionality reduction, and it has gained intensive attention in various fields such as image processing, pattern recognition, and machine learning. However, conventional graph-based dimensionality reduction algorithms usually depend on a fixed weight graph called similarity matrix, which seriously affects the subsequent feature extraction process. In this paper, a novel structured optimal graph based sparse feature extraction (SOGSFE) method for semi-supervised learning is proposed. In the proposed method, the local structure learning, sparse representation, and label propagation are simultaneously framed to perform data dimensionality reduction. In particular the similarity matrix and the projection matrix are obtained by an iterative calculation manner. The experimental results on several public image datasets demonstrate the robustness and effectiveness of the proposed method.

Sensing Matrix Design for Compressive Spectral Imaging via Binary Principal Component Analysis.

Compressive spectral imaging (CSI) is a framework that captures coded-and-multiplexed low-dimensional projections of spectral data-cubes. In general, the sensing process in many CSI architectures is described using binary matrices, so-called sensing/projection matrices, whose elements can be either random or designed. However, some characteristics of the spectral data, such as the ℓ2-norm or the second moment statistics, can be lost when this dimensionality reduction is performed. Similarly, principal component analysis (PCA) is a data dimensionality reduction technique that minimizes the least-squared error between the spectral data and its low-dimensional projection, but preserving its structure or variance. Thus, PCA can be used to guide the CSI acquisition process by designing the binary sensing matrix. Nonetheless, PCA requires to know the spectral image a-priori, and also, its associated projection matrix is not binary, as required by CSI optical architectures. Therefore, in this paper, an algorithm to design CSI sensing matrices by exploiting the structure-preserving property of the PCA projection is proposed. First, a set of compressive measurements obtained with random sensing matrices is used to rapidly estimate the covariance matrix associated with the spectral data. Then, a new sensing matrix is designed by solving a non-convex optimization problem that finds a set of binary vectors that approximate the principal components of the covariance matrix, thus maximizing the explanation of the data variance. Experimental results show an improvement of up to 3 dB in image reconstruction quality, in terms of the peak signal to noise ratio (PSNR), when the binary PCA-based sensing matrices are used and compared with conventional random sensing matrices and state-of-art designed matrices based on PCA.

Supervised and unsupervised machine learning for automated scoring of sleep-wake and cataplexy in a mouse model of narcolepsy.

Despite commercial availability of software to facilitate sleep-wake scoring of electroencephalography (EEG) and electromyography (EMG) in animals, automated scoring of rodent models of abnormal sleep, such as narcolepsy with cataplexy, has remained elusive. We optimize two machine-learning approaches, supervised and unsupervised, for automated scoring of behavioral states in orexin/ataxin-3 transgenic mice, a validated model of narcolepsy type 1, and additionally test them on wild-type mice. The supervised learning approach uses previously labeled data to facilitate training of a classifier for sleep states, whereas the unsupervised approach aims to discover latent structure and similarities in unlabeled data from which sleep stages are inferred. For the supervised approach, we employ a deep convolutional neural network architecture that is trained on expert-labeled segments of wake, non-REM sleep, and REM sleep in EEG/EMG time series data. The resulting trained classifier is then used to infer on the labels of previously unseen data. For the unsupervised approach, we leverage data dimensionality reduction and clustering techniques. Both approaches successfully score EEG/EMG data, achieving mean accuracies of 95% and 91%, respectively, in narcoleptic mice, and accuracies of 93% and 89%, respectively, in wild-type mice. Notably, the supervised approach generalized well on previously unseen data from the same animals on which it was trained but exhibited lower performance on animals not present in the training data due to inter-subject variability. Cataplexy is scored with a sensitivity of 85% and 57% using the supervised and unsupervised approaches, respectively, when compared to manual scoring, and the specificity exceeds 99% in both cases.

Scalable and Adaptive Fuzzy Grid Partitioning for Enhanced Rule Generation in Complex Decision-Making Systems

This research focuses on addressing the challenges of rule generation in complex decision-making systems by proposing a scalable and adaptive fuzzy grid partitioning approach. Traditional rule generation methods often struggle to handle large datasets and dynamic environments, leading to decreased accuracy and computational inefficiencies. In this study, we present a novel approach that integrates scalable and adaptive techniques to enhance the accuracy, efficiency, and interpretability of rule-based frameworks. The scalable fuzzy grid partitioning algorithm efficiently partitions the attribute space, allowing for the generation of rules in decision-making systems with a large number of data points. By incorporating data parallelization and dimensionality reduction techniques, the approach mitigates computational complexity while maintaining rule generation accuracy. Furthermore, the adaptive fuzzy grid partitioning algorithm dynamically adjusts the partitioning structure based on changing conditions, capturing evolving patterns and ensuring the relevancy and reliability of the generated rules over time. The generated rules are evaluated using fuzzy rule evaluation functions, which consider the degree of membership in the corresponding fuzzy grid cells. This evaluation process ranks and selects the rules based on their firing strengths, providing an interpretable decision-making framework for complex systems. The approach enhances the interpretability of the generated rules by capturing the uncertainties and complexities inherent in decision-making processes. To validate the effectiveness of the proposed approach, we conducted experiments using a credit risk assessment case example. The results demonstrate improved accuracy and efficiency compared to traditional rule generation methods. The generated rules offer transparency and insight into the factors influencing credit risk assessments, enabling informed decision-making. However, this research has some limitations, including potential dataset dependencies, the choice of fuzzy membership functions, computational complexity, and the need for further evaluation metrics and real-world implementation considerations. Future research should focus on addressing these limitations and exploring the applicability of the proposed approach in diverse domains. In conclusion, the scalable and adaptive fuzzy grid partitioning approach presented in this research offers a promising solution to the challenges of rule generation in complex decision-making systems. By addressing scalability, adaptability, and interpretability, this approach enhances the accuracy and efficiency of rule-based frameworks, paving the way for more effective decision support systems in various domains.

Mapping impervious surface fractions using automated Fisher transformed unmixing

Spatial explicit monitoring of impervious surface fractions provides essential information for urban planning, urban disaster prevention and mitigation. Moderate spectral and spatial resolution satellite observation has become a tool to perform such monitoring across different spatial and temporal scales. Yet, one major challenge in urban remote sensing is the high spatial heterogeneity causing the presence of image pixels constituting a mixture of different land cover materials. Here we apply an automated unmixing procedure specifically designed to optimally account for the pronounced spectral variability (i.e. high within-class and low between-class variability) prevailing in urban settings. The approach is based on Fisher Discriminant Analysis (FDA), a data dimensionality reduction technique in which between-class variability is maximized and within-class variability is minimized while transforming spectra from the reflectance to the Fisher feature space. We integrated the FDA transformation in the widely-applied Multiple Endmember Spectral Mixture Analysis (MESMA) approach in an attempt to more effectively address the endmember variability problem. We made use of online spectral libraries to train the Fisher transformation parameters and steer the automated selection of image endmembers. Our Fisher transformed MESMA algorithm (F-MESMA) was tested on medium spatial resolution Landsat scenes and impervious fraction maps were generated for five major cities located in different regions of the world. The output of F-MESMA was compared to the output of state-of-the-art spectral mixture analysis approaches which were specifically designed to reduce endmember variability issues. We could demonstrate that compared to other spectral transformation approaches, the ratio of within- vs between-class variability of the endmembers was most strongly reduced after applying the Fisher transformation. As a direct consequence, across all five cities F-MESMA consistently provided the most accurate impervious surface fraction estimates (RMSEF-MESMA = 0.13 vs. RMSEalternative approaches = [0.16–0.17]).

Comprehending the IoT cyber threat landscape: A data dimensionality reduction technique to infer and characterize Internet-scale IoT probing campaigns

The resource-constrained and heterogeneous nature of Internet-of-Things (IoT) devices coupled with the placement of such devices in publicly accessible venues complicate efforts to secure these devices and the networks they are connected to. The Internet-wide deployment of IoT devices also makes it challenging to operate security solutions at strategic locations within the network or to identify orchestrated activities from seemingly independent malicious events from such devices. Therefore, in this paper, we initially seek to determine the magnitude of IoT exploitations by examining more than 1 TB of passive measurement data collected from a/8 network telescope and by correlating it with 400 GB of information from the Shodan service. In the second phase of the study, we conduct in-depth discussions with Internet Service Providers (ISPs) and backbone network operators, as well as leverage geolocation databases to not only attribute such exploitations to their hosting environment (ISPs, countries, etc.) but also to classify such inferred IoT devices based on their hosting sector type (financial, education, manufacturing, etc.) and most abused IoT manufacturers. In the third phase, we automate the task of alerting realms that are determined to be hosting exploited IoT devices. Additionally, to address the problem of inferring orchestrated IoT campaigns by solely observing their activities targeting the network telescope, we further introduce a theoretically sound technique based on L1-norm PCA, and validate the utility of the proposed data dimensionality reduction technique against the conventional L2-norm PCA. Specifically, we identify “in the wild” IoT coordinated probing campaigns that are targeting generic ports and campaigns specifically searching for open resolvers (for amplification purposes). The results reveal more than 120,000 Internet-scale exploited IoT devices, some of which are operating in critical infrastructure sectors such as health and manufacturing. We also infer 140 large-scale IoT-centric probing campaigns; a sample of which includes a worldwide distributed campaign where close to 40% of its population includes video surveillance cameras from a single manufacturer, and another very large inferred coordinated campaign consisting of more than 50,000 IoT devices. The reported findings highlight the insecurity of the IoT paradigm at large and thus demonstrate the importance of understanding such evolving threat landscape.

Комбинированные методы спектрального анализа неоднородного потока зерновых смесей

The problem of detecting Fusarium fungi infected grains and identifying the varieties of the components of grain mixtures based on the combined use of various spectral measurement methods in various wavelength ranges is considered. A method for measuring combined reflectance and transmittance spectra of the elements of non-uniform grain flow is described. The results of the spectral measurement are processed using neural network based classification algorithms combined with data dimensionality reduction techniques. The probabilities of incorrect recognition for various numbers of features and combinations of spectral methods are estimated

Optimal Chemical Grouping and Sorbent Material Design by Data Analysis, Modeling and Dimensionality Reduction Techniques

The ultimate goal of the Texas A&M Superfund program is to develop comprehensive tools and models for addressing exposure to chemical mixtures during environmental emergency-related contamination events. With that goal, we aim to design a framework for optimal grouping of chemical mixtures based on their chemical characteristics and bioactivity properties, and facilitate comparative assessment of their human health impacts through read-across. The optimal clustering of the chemical mixtures guides the selection of sorption material in such a way that the adverse health effects of each group are mitigated. Here, we perform (i) hierarchical clustering of complex substances using chemical and biological data, and (ii) predictive modeling of the sorption activity of broad-acting materials via regression techniques. Dimensionality reduction techniques are also incorporated to further improve the results. We adopt several recent examples of chemical substances of Unknown or Variable composition Complex reaction products and Biological materials (UVCB) as benchmark complex substances, where the grouping of them is optimized by maximizing the Fowlkes-Mallows (FM) index. The effect of clustering method and different visualization techniques are shown to influence the communication of the groupings for read-across.

Using Dimensionality Reduction Techniques for Refining Passive Indoor Positioning Systems Based on Radio Fingerprinting.

Indoor positioning methods based on fingerprinting and radio signals rely on the quality of the radio map. For example, for room-level classification purposes, it is required that the signal observations related to each room exhibit significant differences in their RSSI values. However, it is difficult to verify and visualize that separability since radio maps are constituted by multi-dimensional observations whose dimension is directly related to the number of access points or monitors being employed for localization purposes. In this paper, we propose a refinement cycle for passive indoor positioning systems, which is based on dimensionality reduction techniques, to evaluate the quality of a radio map. By means of these techniques and our own data representation, we have defined two different visualization methods to obtain graphical information about the quality of a particular radio map in terms of overlapping areas and outliers. That information will be useful to determine whether new monitors are required or some existing ones should be moved. We have performed an exhaustive experimental analysis based on a variety of different scenarios, some deployed by our own research group and others corresponding to a well-known existing dataset widely analyzed by the community, in order to validate our proposal. As we will show, among the different combinations of data representation methods and dimensionality reduction techniques that we discuss, we have found that there are some specific configurations that are more useful in order to perform the refinement process.

Using data dimensionality reduction for recognition of incomplete dynamic gestures

Continuous gesture spotting is a major topic in human-robot interaction (HRI) research. Human gestures are captured by sensors that provide large amounts of data that can be redundant or incomplete, correlated or uncorrelated. Data dimensionality reduction (DDR) techniques allow to represent such data in a low-dimensional space, making the classification process more efficient. This study demonstrates that DDR can improve the classification accuracy and allows the classification of gesture patterns with incomplete data, i.e., with the initial 25%, 50% or 75% of data representing a given dynamic gesture (DG) - time series of positional and hand shape data. Re-sampling raw data with bicubic interpolation and principal component analysis (PCA) were used as DDR methods. The performance of different classifiers is compared in the classification 95 different signs of the UCI Australian Sign Language (High Quality) Dataset. Experimental tests indicate that the use of PCA-based features result in a classification accuracy that is higher with 25% of gesture data (93% accuracy) than with 100% of gesture data (82% accuracy). These results were obtained from a non-trained data set and the recognized gestures are used to control a robot in an collaborative process.

Interpretation of FTIR spectra of polymers and Raman spectra of car paints by means of likelihood ratio approach supported by wavelet transform for reducing data dimensionality.

The problem of interpretation of common provenance of the samples within the infrared spectra database of polypropylene samples from car body parts and plastic containers as well as Raman spectra databases of blue solid and metallic automotive paints was under investigation. The research involved statistical tools such as likelihood ratio (LR) approach for expressing the evidential value of observed similarities and differences in the recorded spectra. Since the LR models can be easily proposed for databases described by a few variables, research focused on the problem of spectra dimensionality reduction characterised by more than a thousand variables. The objective of the studies was to combine the chemometric tools easily dealing with multidimensionality with an LR approach. The final variables used for LR models' construction were derived from the discrete wavelet transform (DWT) as a data dimensionality reduction technique supported by methods for variance analysis and corresponded with chemical information, i.e. typical absorption bands for polypropylene and peaks associated with pigments present in the car paints. Univariate and multivariate LR models were proposed, aiming at obtaining more information about the chemical structure of the samples. Their performance was controlled by estimating the levels of false positive and false negative answers and using the empirical cross entropy approach. The results for most of the LR models were satisfactory and enabled solving the stated comparison problems. The results prove that the variables generated from DWT preserve signal characteristic, being a sparse representation of the original signal by keeping its shape and relevant chemical information.