Random Characteristics Research Articles

Recursive Hyperparameter-Free Criterion Learning

In the context of adaptive signal processing for non-Gaussian noise scenarios, the paradigm of information theoretic learning (ITL) has emerged useful due to their incorporation of higher order error-statistics, their improved convergence, and for their motivation from the standpoint of statistical mechanics. However, these ITL criteria are well-known to depend on scenario-dependent hyperparameter choices, whose optimal values, in-turn, depend on scenario dependent noise-statistics. This brief proposes hyperparameter free criterion learning using random Fourier features (RFF), which alleviates hyperparameter-dependence, and allows for scenario-independent generalization for underlying noise-distributions. For the proposed approach, detailed convergence analysis is presented and validated via relevant case-studies.

Thematic Study of Naipaul’s Selected Works

Theme defines a major subject, idea or underlying meaning that is being discussed or a writer explores in a piece of writing. It is the suggested view taken on the main idea or message of any fictional, dramatic or nonfictional story. For instance, love can be the subject but desire for love might be the theme. The setting, characters, plot, dialogue are combined together to convey the theme of any piece of writing. Theme is not only the storyline or description but also something more than it. It is the center of any narrative that runs throughout a plot. Betrayal, love, life and death, courage, good vs. evil, revenge, beauty, family etc. are some of the common themes in literature. Important concepts and messages encountered by the characters and the setting of a tale are communicated through themes. A story without a theme is just a collection of random characters and events. It serves as the component that gives a tale or poem its meaning. This article discusses different kinds of themes used within the works of V.S. Naipaul. The main themes in his writings are escapism and disillusionment, Exile and alienation, displacement or migration, the search of a stable sense of personal identity, rootlessness and many more.

SUDF-RS: A new foreign exchange rate prediction method considering the complementarity of supervised and unsupervised deep representation features

The supervised and unsupervised deep representation features have been adopted and verified their effective roles separately in the foreign exchange rate prediction (FERP). However, the complementarity of these two types of features for FERP remained unexplored. To address this problem, we proposed a novel method for one day ahead FERP, which improves Random Subspace by simultaneously considering both Supervised and Unsupervised Deep representation Features, namely, SUDF-RS. Feature extraction and model construction are two important stages in the SUDF-RS method. Firstly, the supervised and unsupervised deep representation features are extracted by the long short-term memory networks and deep belief networks respectively. Secondly, an improved RS method, which incorporates random forest-based feature weighting mechanism is developed to generate high-quality feature subsets. Thirdly, each feature subset is used to train the corresponding base learner and the final results are obtained by averaging the results of each base learner. Experiments on three exchange rate datasets, namely EUR/USD, GBP/USD and USD/JPY validate that the proposed SUDF-RS significantly outperforms the benchmark methods in terms of both MAPE and RMSE.

Bearing Fault Diagnosis Based on Randomized Fisher Discriminant Analysis.

In this paper, a novel randomized Fisher discriminant analysis (RFDA) based bearing fault diagnosis method is proposed. First, several representative time-domain features are extracted from the raw vibration signals. Second, linear Fisher discriminant analysis (FDA) is extended to nonlinear FDA named RFDA by introducing the random feature map to deal with the non-linearity issue. Specifically, the extracted time-domain features data are mapped onto a high-dimensional space using the random feature map function rather than kernel functions. Third, the time-domain features are fed into the built RFDA model to extract the discriminant features for diagnosis. Moreover, a Bayesian inference is employed to identify the class of the collected vibration signals to diagnose the bearing status. The proposed method uses random Fourier features to approximate the kernel matrix in the kernel Fisher discriminant analysis. Through employing randomized Fisher discriminant analysis, the nonlinearity issue is dealt with, and the computational burden is remarkably reduced compared to the kernel Fisher discriminant analysis (KFDA). To illustrate the superior performance of the proposed RFDA-based bearing fault diagnosis method, comparative experiments are conducted on two widely used datasets, the Case Western Reserve University (CWRU) bearing dataset and the Paderborn University (PU) bearing dataset. For the CWRU dataset, the computation time of RFDA is much shorter than KFDA, while the accuracy rate reaches the same level of KFDA. For the PU dataset, the accuracy rate of RFDA is slightly higher than KFDA, and the computation time is only 44.14% of KFDA.

Abrolhos Magmatic Province petrogenesis and its link with the Vitória-Trindade Ridge, Southeast Brazilian Margin, South Atlantic Ocean

The Abrolhos Magmatic Province (AMP) is an example of an igneous province with about 63,000 km2 located at the Continent-Ocean Boundary (COB), into the Southeast Brazilian Margin. The AMP emerges as five small islands (Santa Bárbara, Redonda, Siriba, Sueste, and Guarita) which integrate the Abrolhos Archipelago located about 55 km offshore Brazil. The studied rocks belong to a transitional basalt series of alkaline affinity, with evolved rocks with high TiO2 contents. This work presents new detailed fieldwork mapping, petrographic, and whole-rock chemistry data, besides Sr–Nd isotopic compositions from AMP rocks. Mapped magmatic rocks in the Abrolhos Islands have been described as extrusive rocks, but we point out that they are shallow intrusions, mostly sills, and should be grouped into diabase units based on fieldwork and textural features. The AMP lithogeochemical and isotopic data suggest the operation of differentiation processes more complex than simple fractional crystallization or AFC. Our model mixing calculations based on new and compiled Sr-Nd-Pb isotopic data suggest a mixture with a prevailing depleted asthenosphere mantle source (represented by DMM; ca. 80%) metasomatized by an enriched mantle I (EMI) component (≤ 10%) and a possible HIMU-type constituent (up to 10%) in the AMP genesis. The assimilation of subducted slabs of the oceanic crust associated with the HIMU signatures is possibly linked to the Brasiliano Event due to the range of the AMP Nd TDM model ages, from 407 to 767 Ma. A viable mechanism for the EMI end-member contribution could either be a physical detachment of the South American subcontinental lithospheric mantle during the breakup of the Gondwana or lithospheric delamination of the South American plate caused by an edge-driven convection mechanism. The AMP and the Vitória-Trindade Ridge (VTR) show an eastward decreasing age pattern from the older ca. 60 Ma Abrolhos Province to the younger Martin Vaz and Trindade Islands (ca. 4 – 0.1 Ma), located ca. 1200 km away from the Brazilian coastline. The volcanic alignment between the VTR and AMP, along with the overlap of geochemical and isotopic data of their different igneous rocks, cannot be a random feature but instead represent the sampling of similar shallow mantle reservoirs, thus suggesting a cogenetic relationship. Finally, a possible petrogenetic link between the AMP and VTR magmatism is discussed.

Integrated wavelet-learning method for macroscopic mechanical properties prediction of concrete composites with hierarchical random configurations

In this research, a novel integrated wavelet-learning approach with high-accuracy and efficient prediction capability is developed to predict the effective elastic properties of concrete composites with hierarchical random configurations and heterogeneous elasticity in multiple spatial scales. The significant contributions of this work include: (1) The random structural heterogeneities and material uncertainties of concrete composites in meso-scale and micro-scale are modeled on the basis of modified authors’ previous researches and Weibull probabilistic model separately. (2) Innovative stochastic three-scale asymptotic homogenization method (STAHM), background grid technique and padding strategy are presented to thoroughly acquire the doubly random geometry and material features of concrete composites with hierarchical random configurations for constructing concrete material databases. (3) The wavelet transform is utilized to preprocess the high-dimensional data features of the material databases, and the wavelet coefficients are employed as the novel input neurons for efficiently establishing more plain wavelet-neural networks that possess increasing capacity and resisting noise for approximating mappings of high complexity and nonlinearity. The numerical experiments and achieved results demonstrate that the integrated wavelet-learning approach is robust for the high-accuracy and efficient prediction of effective elastic properties of concrete materials with random heterogeneity patterns in multiple spatial scales.

A DC Arc Fault Detection Method Based on AR Model for Photovoltaic Systems

DC arc faults are dangerous to photovoltaic (PV) systems and can cause serious electric fire hazards and property damage. Because the PV inverter works in a high−frequency pulse width modulation (PWM) control mode, the arc fault detection is prone to nuisance tripping due to PV inverter noises. An arc fault detection method based on the autoregressive (AR) model is proposed. A test platform collects the database of this research according to the UL1699B standard, in which three different types of PV inverters are taken into consideration to make it more generalized. The arc current can be considered a nonstationary random signal while the noise of the PV inverter is not. According to the difference in randomness features between an arc and the noise, a detection method based on the AR model is proposed. The Burg algorithm is used to determine model coefficients, while the Akaike Information Criterion (AIC) is applied to explore the best order of the proposed model. The correlation coefficient difference of the model coefficients plays a role as a criterion to identify if there is an arc fault. Moreover, a prototype circuit based on the TMS320F28033 MCU is built for algorithm verification. Test results show that the proposed algorithm can identify an arc fault without a false positive under different PV inverter conditions. The fault clearing time is between 60 ms to 80 ms, which can meet the requirement of 200 ms specified by the standard.

Random Interval Tracking System for Virtual Learning

With the advent of computer technology, E-Learning has become an increasingly popular learning approach in higher Education institutions. Due to the outbreak of COVID-19 pandemic that demands substantial modification in the teaching-learning process across the globe. However, the student’s attendance management in virtual classroom is quite difficult. In this paper I introduce a smart intelligent virtual classroom, which include ML based face recognition system. Further, it can generate dedicated attendance reports, pinpointing students’ attention during virtual learning at random time intervals. Moreover, the novel random interval tracking system can also prevent the dropping out of participants from the virtual classroom. The distinctive feature of randomness in Random Interval Tracking System (RITS) ensures that student’s attention and engagement in virtual classroom are enhanced.

Discovering biomarkers of hepatocellular carcinoma from single-cell RNA sequencing data by cooperative games on gene regulatory network

Effective and reliable biomarker is a promising means to achieve early diagnosis of cancer. The availability of high-throughput single-cell RNA sequencing (scRNA-seq) data opens an unprecedented opportunity of discovering biomarkers by developing machine learning and feature selection methods. At present, the existing biomarker screening methods, such as recursive feature elimination (RFE), often treat genes as isolated features, ignoring their embedding complex network relationship. In addition, the interpretability of a cancer biomarker discovery model is as important as its classification accuracy. To address these problems, we propose a game theoretic method to discover gene modules serving as biomarkers on gene regulatory network (GRN) that can better distinguish hepatocellular carcinoma (HCC) samples with healthy ones. Specifically, the network-based game theory method, called NGTM, is an interpretable module exploration of supervised feature selection procedure. We regard the process of gene-to-model selection as a cooperative game. The contribution of each feature in combination is evaluated by cooperative game theoretic metrics, that is, Shapley values. The extension strategy of gene module is conducted on GRN in the form of subnetwork, and NGTM makes the biomarker recognition easily interpretable. Furthermore, our method is statistically verified by Akaike information criterion (AIC) in model selection. There is a strong correlation between AIC and the area under curve in classification. In comparison study, we test the wrapper RFE and random feature extraction methods on random forest under the same conditions. NGTM achieves relatively better classification performances which prove its advantage. The enriched dysfunctions in biomarkers are also consistent with prior knowledge of the occurrence and development of HCC.

Predictive models demonstrate age-dependent association of subcortical volumes and cognitive measures.

Whether brain matter volume is correlated with cognitive functioning and higher intelligence is controversial. We explored this relationship by analysis of data collected on 193 healthy young and older adults through the "Leipzig Study for Mind-Body-Emotion Interactions" (LEMON) study. Our analysis involved four cognitive measures: fluid intelligence, crystallized intelligence, cognitive flexibility, and working memory. Brain subregion volumes were determined by magnetic resonance imaging. We normalized each subregion volume to the estimated total intracranial volume and conducted training simulations to compare the predictive power of normalized volumes of large regions of the brain (i.e., gray matter, cortical white matter, and cerebrospinal fluid), normalized subcortical volumes, and combined normalized volumes of large brain regions and normalized subcortical volumes. Statistical tests showed significant differences in the performance accuracy and feature importance of the subregion volumes in predicting cognitive skills for young and older adults. Random forest feature selection analysis showed that cortical white matter was the key feature in predicting fluid intelligence in both young and older adults. In young adults, crystallized intelligence was best predicted by caudate nucleus, thalamus, pallidum, and nucleus accumbens volumes, whereas putamen, amygdala, nucleus accumbens, and hippocampus volumes were selected for older adults. Cognitive flexibility was best predicted by the caudate, nucleus accumbens, and hippocampus in young adults and caudate and amygdala in older adults. Finally, working memory was best predicted by the putamen, pallidum, and nucleus accumbens in the younger group, whereas amygdala and hippocampus volumes were predictive in the older group. Thus, machine learning predictive models demonstrated an age-dependent association between subcortical volumes and cognitive measures. These approaches may be useful in predicting the likelihood of age-related cognitive decline and in testing of approaches for targeted improvement of cognitive functioning in older adults.

Structured random receptive fields enable informative sensory encodings.

Brains must represent the outside world so that animals survive and thrive. In early sensory systems, neural populations have diverse receptive fields structured to detect important features in inputs, yet significant variability has been ignored in classical models of sensory neurons. We model neuronal receptive fields as random, variable samples from parameterized distributions and demonstrate this model in two sensory modalities using data from insect mechanosensors and mammalian primary visual cortex. Our approach leads to a significant theoretical connection between the foundational concepts of receptive fields and random features, a leading theory for understanding artificial neural networks. The modeled neurons perform a randomized wavelet transform on inputs, which removes high frequency noise and boosts the signal. Further, these random feature neurons enable learning from fewer training samples and with smaller networks in artificial tasks. This structured random model of receptive fields provides a unifying, mathematically tractable framework to understand sensory encodings across both spatial and temporal domains.

A novel short-term electrical load forecasting framework with intelligent feature engineering

The emergence of many new types of electrical loads, such as security systems and charging piles, has resulted in a diversified trend of building load-end components. This trend is coupled with chaotic influencing factors with time-scale differences such as multiple periods and special events, thereby causing difficulties in extracting multiple time-scale features (including regular electricity behavior features in a long term and random electricity behavior features in an ultrashort term) and reducing the quality of short-term forecasting. To this end, a two-level intelligent feature engineering (IFE) and serial multi-timescale forecasting framework is designed in this work. First, the first-level IFE is proposed to deal with weather and calendar effects and to extract regular electricity behavior features in a long term, thus helping the 24-h ahead forecasting model to obtain a stable forecast value of the load trend. Second, based on the day-ahead trend component forecasting (DATCF) value obtained in the previous step, the second-level IFE is proposed to extract the random electricity behavior features in an ultrashort term. The aim is to improve the forecasting capability of the 15-min ahead forecasting model for load fluctuation detail. This work eliminates redundant and irrelevant features through a two-level IFE and balances the stability for load trend forecasting and the accuracy for load fluctuation detail forecasting through a serial multi-timescale forecasting framework. Finally, the simulation results indicate that the root mean square error (RMSE) of the proposed method is less than that of the control group by 8.98–31.56 kW, and the RMSE of the proposed method decreased by 1.27–66.13 kW during peak hours with strong volatility.

Application of Motion Effect Evaluation Algorithm Based on Random Forest.

The development of big data technology makes the feature selection technology gradually perfect. The advantages of different feature selection technologies are different. Among them, random forest algorithm belongs to integrated feature selection algorithm. Integrated processing of classification results can screen out the most representative feature impact. Based on this background and random forest algorithm, this paper analyzes the evaluation of motion effect. After the measurement, this paper obtains the body data before and after the training. After the calculation, the change data of the body index are determined. The random forest feature selection method is used as the carrier to determine the corresponding index attribute set. In the process of data set input, the corresponding whole input data set is formed through data classification. The completion of training, through the comparative experiment, is conducive to clear the degree of influence of physical indicators and then complete the exercise effect evaluation. The research shows that the random forest algorithm has significant advantages in the evaluation of sports effect and can effectively improve the accuracy of classification.

Adversarial examples for network intrusion detection systems

Machine learning-based network intrusion detection systems have demonstrated state-of-the-art accuracy in flagging malicious traffic. However, machine learning has been shown to be vulnerable to adversarial examples, particularly in domains such as image recognition. In many threat models, the adversary exploits the unconstrained nature of images–the adversary is free to select some arbitrary amount of pixels to perturb. However, it is not clear how these attacks translate to domains such as network intrusion detection as they contain domain constraints, which limit which and how features can be modified by the adversary. In this paper, we explore whether the constrained nature of networks offers additional robustness against adversarial examples versus the unconstrained nature of images. We do this by creating two algorithms: (1) the Adapative-JSMA, an augmented version of the popular JSMA which obeys domain constraints, and (2) the Histogram Sketch Generation which generates adversarial sketches: targeted universal perturbation vectors that encode feature saliency within the envelope of domain constraints. To assess how these algorithms perform, we evaluate them in a constrained network intrusion detection setting and an unconstrained image recognition setting. The results show that our approaches generate misclassification rates in network intrusion detection applications that were comparable to those of image recognition applications (greater than 95%). Our investigation shows that the constrained attack surface exposed by network intrusion detection systems is still sufficiently large to craft successful adversarial examples – and thus, network constraints do not appear to add robustness against adversarial examples. Indeed, even if a defender constrains an adversary to as little as five random features, generating adversarial examples is still possible.

Continuous Encryption Functions for Security Over Networks

This paper presents a study of continuous encryption functions (CEFs) of secret feature vectors for security over networks, which include physical layer encryption for wireless communications and biometric template security for online Internet applications. While CEFs are defined here to include all prior continuous one-way functions, a good CEF is defined to be a continuous function that turns a random feature vector of limited dimension into a long sequence of numbers in such a way that it is hard to invert and hard to substitute, it has no or little amplification of noise, and its output samples have zero or near-zero correlations and have identical or nearly identical distributions. A number of prior CEFs, such as dynamic random projection, index-of-max hashing and higher-order polynomials, are all shown to fail on these criteria. Based on selected components of singular value decomposition (SVD) of randomly modulated matrices of the feature vector, a family of SVD-CEFs is proposed. Such a SVD-CEF is shown to meet all the criteria for a good CEF and outperform the prior CEFs significantly.

Random Features for Kernel Approximation: A Survey on Algorithms, Theory, and Beyond.

The class of random features is one of the most popular techniques to speed up kernel methods in large-scale problems. Related works have been recognized by the NeurIPS Test-of-Time award in 2017 and the ICML Best Paper Finalist in 2019. The body of work on random features has grown rapidly, and hence it is desirable to have a comprehensive overview on this topic explaining the connections among various algorithms and theoretical results. In this survey, we systematically review the work on random features from the past ten years. First, the motivations, characteristics and contributions of representative random features based algorithms are summarized according to their sampling schemes, learning procedures, variance reduction properties and how they exploit training data. Second, we review theoretical results that center around the following key question: how many random features are needed to ensure a high approximation quality or no loss in the empirical/expected risks of the learned estimator. Third, we provide a comprehensive evaluation of popular random features based algorithms on several large-scale benchmark datasets and discuss their approximation quality and prediction performance for classification. Last, we discuss the relationship between random features and modern over-parameterized deep neural networks (DNNs), including the use of high dimensional random features in the analysis of DNNs as well as the gaps between current theoretical and empirical results. This survey may serve as a gentle introduction to this topic, and as a users' guide for practitioners interested in applying the representative algorithms and understanding theoretical results under various technical assumptions. We hope that this survey will facilitate discussion on the open problems in this topic, and more importantly, shed light on future research directions. Due to the page limit, we suggest the readers refer to the full version of this survey https://arxiv.org/abs/2004.11154.

A Data-Driven Short-Term PV Generation and Load Forecasting Approach for Microgrid Applications

The data-driven (DD) is a systematic approach to improve the data and model by deriving/adding features to address the problem identified during the iterative loop of forecasting model development. This article proposes a DD framework for forecasting short-term PV generation and load demand. A framework of three stages with a unique contribution in each stage, such as generalizing data preprocessing steps (stage-1), multivariate feature generation and selection (stage-2), and model hyperparameter tuning (stage-3) for further improvement in forecasting is proposed. It focuses on data as well as forecasting models. The whole process is analyzed using the time-series measured data collected from a real-life demonstration project in Ireland. Data preprocessing is generalized for both generation and demand forecasting under the same framework. The relevant features are selected with the help of the proposed random forest sequential forward feature selection algorithm. Hyperparameters are tuned through tree-structured Parzen estimator algorithm for further improvement. In addition, the performance of the classical autoregressive integrated moving average model is compared with the machine learning-based gate recurrent unit, long short term memory, recurrent neural network, and convolutional neural network models. Results show that the data-driven forecasting model framework systematically improves the model performance. The seasonal variation has also a high impact on the model performances.

Resource Allocation and Slicing Puncture in Cellular Networks With eMBB and URLLC Terminals Coexistence

Ultrareliable low-latency communication (URLLC) and enhanced mobile broadband (eMBB) are two types of services with delay Quality-of-Service (QoS) demands. Considering the random and sporadic URLLC packets arrival feature, slicing puncture is believed to be the suitable method to support coexistence communication scenario of eMBB and URLLC terminals. However, slicing puncture in a short time is not trivial, and needs accurate wireless resource scheduling. At the same time, it may damage the QoS of eMBB service. All of these make the QoS guaranteed scheduling problem more challenging. In this article, we investigate the bandwidth, power allocations, slice puncturing problems to find the way satisfying services’ individual QoS demands. For the scheduling of eMBB service, we formulate a joint optimization problem for bandwidth and power allocations with long-term constraints of queues backlog. To solve this problem, we utilize the Lyapunov drift-plus-penalty (DPP) method to establish the relationship between the long-term constraints and the short-term optimization problem, which means that the long-term constraints are gradually satisfied by the proposed strategy at each step. We further divide the short-term optimization problem into two subproblems and adopt a block coordinate descent (BCD) algorithm to reduce the computation complexity. Then, we put forward a one-to-one matching method to solve the integer programming in resource block (RB) allocation and slicing puncture problems. Numerical results demonstrate that the proposed dynamic resource allocation and puncturing strategy (DRAPS) can solve the scheduling problem of eMBB and URLLC services in the presence of multiple randomness of channel-state information (CSI) and URLLC packet arrival.

Approximate kernel PCA: Computational versus statistical trade-off

Kernel methods are powerful learning methodologies that allow to perform nonlinear data analysis. Despite their popularity, they suffer from poor scalability in big data scenarios. Various approximation methods, including random feature approximation, have been proposed to alleviate the problem. However, the statistical consistency of most of these approximate kernel methods is not well understood except for kernel ridge regression wherein it has been shown that the random feature approximation is not only computationally efficient but also statistically consistent with a minimax optimal rate of convergence. In this paper, we investigate the efficacy of random feature approximation in the context of kernel principal component analysis (KPCA) by studying the trade-off between computational and statistical behaviors of approximate KPCA. We show that the approximate KPCA is both computationally and statistically efficient compared to KPCA in terms of the error associated with reconstructing a kernel function based on its projection onto the corresponding eigenspaces. The analysis hinges on Bernstein-type inequalities for the operator and Hilbert–Schmidt norms of a self-adjoint Hilbert–Schmidt operator-valued U-statistics, which are of independent interest.

Sequential feature selection for heart disease detection using random forest

Heart disease identification is one of the most challenging task that requires highly experienced cardiologists. However, in developing nations such as Ethiopia, there are a few cardiologists and heart disease detection is more challenging. As an alternative solution to cardiologist, this study proposed a more effective model for heart disease detection by employing random forest and sequential feature selection (SFS). SFS is an effective approach to improve the performance of random forest model on heart disease detection. SFS removes unrelated features in heart disease dataset that tends to mislead random forest model on heart disease detection. Thus, removing inappropriate and duplicate features from the training set with sequential feature selection approach plays significant role in improving the performance of the proposed model. The proposed feature selection approach is evaluated using real world clinical heart disease dataset collected from University of California Irvine (UCI) data repository. Empirical test on validation set reveals that the proposed model performs well as compared to the existing methods. Overall, the state of-the-art heart disease detection model with classification accuracy of 98.53% is proposed for heart disease detection using SFS and random forest model.