Congenital heart disease affects approximately 1% of children worldwide, with a number of cases in resource-limited settings remaining undiagnosed through school age. While cardiac auscultation is a key screening method, its effectiveness varies widely, depending on practitioner expertise. This study introduces an innovative artificial intelligence (AI) approach combining conventional machine learning and deep learning techniques to improve heart murmur detection in pediatric populations. By developing an integrated Residual-Recurrent Neural Networks model and analyzing heart sound recordings from 500 pediatric participants, we achieved remarkable diagnostic performance in real-world pediatric clinical settings. At the single recording-level, the model achieved an accuracy of 88.5%, sensitivity of 85.5%, and specificity of 90.7%. Performance improved at the participant-level, with an accuracy of 90.0%, sensitivity of 88.8%, and specificity of 91.2%. The model showed particularly strong results when tested against the PhysioNet database (accuracy 95.2%, sensitivity 91.6%, and specificity 99.1%). This research provides a compelling proof-of-concept for AI-assisted cardiac screening, potentially revolutionizing early detection strategies in pediatric cardiac diseases.

This study proposes a non-parametric ICA method, called ECOPICA, which describes the joint distribution of data by empirical copulas and measures the dependence between recovery signals by an independent test statistic. We employ the grasshopper algorithm to optimize the proposed objective function. Several acceleration tricks are further designed to enhance the computational efficiency of the proposed algorithm under the parallel computing framework. Our simulation and empirical analysis show that ECOPICA produces better and more robust recovery performances than other well-known ICA approaches for various source distribution shapes, especially when the source distribution is skewed or near-Gaussian.

This study considers monitoring photochemical pollutants for anomaly detection based on symbolic interval-valued data analysis. For this task, we construct control charts based on the principal component scores of symbolic interval-valued data. Herein, the symbolic interval-valued data are assumed to follow a normal distribution, and an approximate expectation formula of order statistics from the normal distribution is used in the univariate case to estimate the mean and variance via the method of moments. In addition, we consider the bivariate case wherein we use the maximum likelihood estimator calculated from the likelihood function derived under a bivariate copula. We also establish the procedures for the statistical control chart based on the univariate and bivariate interval-valued variables, and the procedures are potentially extendable to higher dimensional cases. Monte Carlo simulations and real data analysis using photochemical pollutants confirm the validity of the proposed method. The results particularly show the superiority over the conventional method that uses the averages to identify the date on which the abnormal maximum occurred.

SVR-ARMA-GARCH models provide flexible model fitting and good predictive powers for nonlinear heteroscedastic time series datasets. In this study, we explore the change point detection problem in the SVR-ARMA-GARCH model using the residual-based CUSUM test. For this task, we propose an alternating recursive estimation (ARE) method to improve the estimation accuracy of residuals. Moreover, we suggest using a new testing method with a time-varying control limit that significantly improves the detection power of the CUSUM test. Our numerical analysis exhibits the merits of the proposed methods in SVR-ARMA-GARCH models. A real data example is also conducted using BDI data for illustration, which also confirms the validity of our methods.

This paper proposes a KC Score to measure feature importance in clustering analysis of high-dimensional data. The KC Score evaluates the contribution of features based on the correlation between the original features and the reconstructed features in the low dimensional latent space. A KC Score-based feature selection strategy is further developed for clustering analysis. We investigate the performance of the proposed strategy by conducting a study of four single-cell RNA sequencing (scRNA-seq) datasets. The results show that our strategy effectively selects important features for clustering. In particular, in three datasets, our proposed strategy selected less than 5% of the features and achieved the same or better clustering performance than when using all of the features.

In this study, we propose a pattern matching procedure to seize similar price movements of two stocks. First, the algorithm of searching the longest common subsequence is introduced to sieve out the time periods in which the two stocks have the same integrated volatility levels and price rise/drop trends. Next we transform the price data in the found matching time periods to the Bollinger Percent b data. The low frequency power spectra of the transformed data are used to extract trends. Pearson’s chi square test is used to assess similarity of the price movement patterns in the matching periods. Simulation results show the proposed procedure can effectively detect the co-movement periods of two price sequences. Finally, we apply the proposed procedure to empirical high frequency transaction data of NYSE.

Thanks to advanced technologies, ultra-high-frequency limit order book (LOB) data are now available to data analysts. An LOB contains comprehensive information on all transactions in a market. We use LOB data to investigate the high-frequency dynamics of market supply and demand (S-D) and inspect their impacts on intra-daily market trends. The intra-daily S-D curves are fitted with B-spline basis functions. Technique of multi-resolution is introduced to capture inhomogeneous curvature of the S-D curves and a lasso-type criterion is employed to select a common basis set. Based on empirical evidence, we model the time varying coefficients in the B-spline interpolation by vector autoregressive models of order . The Xgboost algorithm is employed to extract information from the areas under the S-D curves to predict the intra-daily market trends. In the empirical study, we analyze the LOB data from LOBSTER (). The results show that the proposed approach is able to recover the S-D curves and has satisfactory performance on both curve and market trend predictions.

Although existing literature on high-dimensional regression models is rich, the vast majority of studies have focused on independent and homogeneous error terms. In this article, we consider the problem of selecting high-dimensional regression models with heteroscedastic and time series errors, which have broad applications in economics, quantitative finance, environmental science, and many other fields. The error term in our model is the product of two components: one time series component, allowing for a short-memory, long-memory, or conditional heteroscedasticity effect, and a high-dimensional dispersion function accounting for exogenous heteroscedasticity. By making use of the orthogonal greedy algorithm and the high-dimensional information criterion, we propose a new model selection procedure that consistently chooses the relevant variables in both the regression and the dispersion functions. The finite sample performance of the proposed procedure is also illustrated via simulations and real data analysis.

In principal component analysis (PCA), principal components are identified by maximizing the component score variance around the mean. However, a practitioner might be interested in capturing the variation in the tail rather than the center of a distribution to, for example, identify the major pollutants from air pollution data. To address this problem, we introduce a new method called Huber-type principal expectile component (HPEC) analysis that uses an asymmetric Huber norm to provide a kind of robust-tail PCA. The statistical properties of HPECs are derived, and a derivative-free optimization approach called particle swarm optimization (PSO) is used to identify HPECs numerically. As a demonstration, HPEC analysis is applied to real and simulated data with encouraging results.

A through-silicon via (TSV) is a conducting copper nail, which provides an electrical connection through a substrate, and is expected to be used extensively to provide highspeed interconnects between the top and bottom of an active die. However, some TSV structural defects such as pinholes and voids are difficult to capture as they commonly affect TSV performance parameters rather than TSV logical function. In order to electrically detect failures, it is necessary to study and analyze electrical characteristics of defects in advance. Testing TSV interconnects for manufacturing defects poses major challenges, and new design-for-test techniques are needed. Here, a novel nondestructive defect detection method using machine learning (ML) is proposed in order to detect void, short, and open defects in TSV-based 3-D ICs. A supervised ML approach is used to build a classification model from training S-parameter data sets containing the defected TSV and the normal TSV. The performance of the random forest classifier is tested for various amounts of void, short, and open defects in TSV-based 3-D-stacked ICs with satisfactory results.

In this study, we use a vector functional autoregressive model to analyze the supply and demand curves of a Limit Order Book (LOB) simultaneously. The supply and demand curves are represented by relevant multi-resolution B-spline basis functions. The corresponding coefficients of the basis functions are shown to follow a vector autoregressive model, which can be applied to the predictions of future demand and supply curves. By computing the areas under the predicted supply and demand curves, we propose a trading strategy based on the LOB data. An empirical study is conducted with the AAPL, MSFT and SPY stocks. Numerical results indicate that the proposed trading strategy has satisfactory investment performance.

Through silicon vias (TSVs) act as media to transport power supply and signals among stacks of three-dimensional stacked integrated circuits (3D-SICs). The testing of TSVs is important as a single faulty TSV can cause an entire stack to fail. There are some TSV structural defects such as pin-holes and voids, which are difficult to capture as they commonly affect TSV performance parameter rather than TSV logical function. In this paper, a novel non-destructive failure detection method using S-parameter is proposed in order to differentiate failure types and also to localize the failure in TSV based 3DS-ICs. In order to electrically detect failures, it is necessary to study and analyze electrical characteristics of failures in advance. In this paper, modeling and analysis of TSV void defects is conducted with S-parameter simulation. Based on machine learning (ML) approach, a void defect detection method for the characterization of voids in TSV structures is proposed by comparing S parameter from the faulty TSV and the fault-free TSV.

Stacked die package using capacitive-based chip-to-chip signaling offers low energy-per-bit costs and high I/O density. Many previously published works on capacitive proximity I/O have focused on mechanical methods for accurate chip alignment. This paper explores design considerations to address stacked die structures with differential signaling chip-to-chip communication based on capacitive coupling. The signal and noise properties of capacitively coupled channels are studied using a 3-D electromagnetic field solver to extract mixed-mode scattering parameters of the differential signal pads. We present side differential and corner differential arrangements and compare their ability to reject noise. We then formulate a functional data analysis approach for assessing the signal transmission quality. Our results indicate that the differential transmitter and receiver pads arrangement and placement can significantly affect the interaction between a channel and its nearest neighbors. The preferred choice for I/O pad placement design depends on the placement arrangements and the overlap ratio of the differential I/O pads.

Estimation of the integrated volatility is an important problem in high-frequency financial data analysis. In this study, we propose a quadratic unbiased estimator of the integrated volatility for stochastic volatility models with microstructure noise. The proposed estimator minimizes the finite sample variance in the class of quadratic estimators based on symmetric Toeplitz matrices. We show the proposed estimator has an asymptotic mixed normal distribution with optimal convergence rate n(-1/4) and achieves the maximum likelihood estimator efficiency for constant volatility case. Simulation results show that our proposed estimator attains better finite sample efficiency than state-of-the-art methods. Finally, a real data analysis is conducted for illustration.

In this paper, we propose a goodness of fit test for continuous time stochastic volatility models based on discretely sampled observations. The proposed test is constructed by measuring deviations between the empirical and true characteristic functions obtained from the hypothesized stochastic volatility model. In this study, both the test statistics based on the fixed and decreasing sampling schemes are taken into consideration. It is shown that under the null, the proposed tests asymptotically follow a weighted sum of products of centered normal random variables. In order to evaluate the proposed tests, a simulation study is performed, in which a bootstrap method is also considered. Finally, a real data analysis is conducted for illustration.

This work aims at estimating inverse autocovariance matrices of long memory processes admitting a linear representation. A modified Cholesky decomposition is used in conjunction with an increasing order autoregressive model to achieve this goal. The spectral norm consistency of the proposed estimate is established. We then extend this result to linear regression models with long-memory time series errors. In particular, we show that when the objective is to consistently estimate the inverse autocovariance matrix of the error process, the same approach still works well if the estimated (by least squares) errors are used in place of the unobservable ones. Applications of this result to estimating unknown parameters in the aforementioned regression model are also given. Finally, a simulation study is performed to illustrate our theoretical findings.

In the literature, traders are often classified into informed and uninformed and the trades from informed traders have market impacts. We investigate these trades by first establishing a scheme to identify the influential trades from the ordinary trades under certain criteria. The differential properties between these two types of trades are examined via the four transaction states classified by the trade price, trade volume, quotes, and quoted depth. Marginal distribution of the four states and the transition probability between different states are shown to be distinct for informed trades and ordinary liquidity trades. Furthermore, four market reaction factors are introduced and logistic regression models of the influential trades are established based on these four factors. Empirical study on the high-frequency transaction data from the NYSE TAQ database show supportive evidence for high correct classification rates of the logistic regression models.

Stochastic differential equation (SDE) models are useful in describing complex dynamical systems in science and engineering. In this study, we consider a monitoring procedure for an early detection of dispersion parameter change in SDE models. The proposed scheme provides a useful diagnostic analysis for phase I retrospective study and develops a flexible and effective control chart for phase II prospective monitoring. A standardized control chart is constructed, and a bootstrap method is used to estimate the mean and variance of the monitoring statistic. The control limit is obtained as an upper percentile of the maximum value of a standard Wiener process. The proposed procedure appears to have a manageable computational complexity for online implementation and also to be effective in detecting changes. We also investigate the performance of the exponentially weighted mean squared control charts for the continuous SDE processes. A simulation method is used to study the empirical sizes and the average run length characteristics of the proposed scheme, which also demonstrates the effectiveness of our method. Finally, we provide an empirical example for illustration. Copyright © 2014 John Wiley & Sons, Ltd.

This paper proposes a new statistic to conduct cross-sectional independence test for the residuals involved in a parametric panel data model. The provided test statistic, which is called linear spectral statistic (LSS), is established based on the characteristic function of the empirical spectral distribution (ESD) of the sample correlation matrix of the residuals. The main advantage of the proposed test statistic is that it can capture nonlinear cross-sectional dependence. Asymptotic theory for a general class of linear spectral statistics is established, as the cross-sectional dimension $N$ and time length $T$ go to infinity proportionally. This type of statistics cover many classical statistics, including the bias-corrected Lagrange Multiplier (LM) test statistic and the likelihood ratio test statistic. Furthermore, the power under a local alternative hypothesis is analyzed and the asymptotic distribution of the proposed statistic under this local hypothesis is also established. Finite sample performance shows that the proposed test statistic works well numerically in each individual case and it can also distinguish some dependent but uncorrelated structures, for example, nonlinear MA(1) models and multiple ARCH(1) models.

Independent component analysis (ICA) is a modern computational method developed in the last two decades. The main goal of ICA is to recover the original independent variables by linear transformations of the observations. In this study, a copula-based method, called COPICA, is proposed to solve the ICA problem. The proposed COPICA method is a semiparametric approach, the marginals are estimated by nonparametric empirical distributions and the joint distributions are modeled by parametric copula functions. The COPICA method utilizes the estimated copula parameter as a dependence measure to search the optimal rotation matrix that achieves the ICA goal. Both simulation and empirical studies are performed to compare the COPICA method with the state-of-art methods of ICA. The results indicate that the COPICA attains higher signal-to-noise ratio (SNR) than several other ICA methods in recovering signals. In particular, the COPICA usually leads to higher SNRs than FastICA for near-Gaussian-tailed sources and is competitive with a nonparametric ICA method for two dimensional sources. For higher dimensional ICA problem, the advantage of using the COPICA is its less storage and less computational effort.