Long-term Correlation Research Articles

Ocean Temperatures Do Not Account for a Record-Setting Winter in the U.S. West

The record-setting winter of 2022–2023 came as an answer to both figurative and literal prayers for political leaders, policy makers, and water managers reliant on snowpacks in the Upper Colorado River Basin, a vital source of water for tens of millions of people across the Western United States. But this “drought-busting” winter was not well-predicted, in part because while interannual patterns of tropical ocean temperatures have a well-known relationship to precipitation patterns across much of the American West, the Upper Colorado is part of a liminal region where these connections tend to be comparatively weak. Using historical sea surface temperature and snowpack records, and leveraging a long-term cross-basin relationship to extend the timeline for evaluation, this analysis demonstrates that the 2022–2023 winter did not present in accordance with other high-snowpack winters in this region, and that the associative pattern of surface temperatures in the tropical Pacific, and snow water equivalent in the regions that stored and supplied most of the water to the Colorado River during the 2022–2023 winter, was not substantially different from a historically incoherent arrangement of long-term correlation. These findings suggest that stochastic variability plays an outsized role in influencing water availability in this region, even in extreme years, reinforcing the importance of other trends to inform water policy and management.

Breast Cancer Image Classification based on Adaptive Interpolation Approach Using Clinical Dataset

In the healthcare and bioinformatics disciplines, the categorization of breast cancer has become an emerging paradigm due to the second most common cause of cancer-related mortality in women. A biopsy is a procedure in which tissue has been examined to determine whether or not it is breast cancer through histopathologists that may lead to a mistaken diagnosis. The research is mainly focused on the patient data (884 case reports of patients) is acquired from the American Oncology Institute implemented with preprocessing techniques consists of missing values and those values are recovered with Novel Modified Interpolation (MI) Method. Deep learning networks effectively detect and assess the pattern for annotating histological data based on the labelling, which preserves time, system cost and enhances the system accuracy. This framework addressed feature acquisition and missing analysis strategies based on entropy confidence weight factor. First, the iterative patterns have been treated as a potential diagnostic rule, and the attention-based rule combination formulates the classification issue based on integrating convolutional and recurrent neural networks, and the short-term and long-term spatial correlations between patches. Second, the key part of label construction is carried out with an entropy confidence-weight factor assessment which detects and predict different patterns to construct the rule for classification. Third, optimization of clustering data by assessing missing parameters based on mean square error and the concept of interpolation to reduce data loss by around 20% and enhance the system accuracy. Simulation results show that the proposed system achieves 91.3% accuracy to state-of-art approaches, potentially allied in clinical applications. Modified Interpolation (MI) method recovered missing values with least mean square error and less data loss of 0.0123 and 1.38% respectively. This method also compared with existing Linear Interpolation (LI) method which is able to recover least men square error and data loss of 3.295 and 18.925% respectively. Comparatively the modified interpolation method recovered the missing values with less mean square error and less data loss.

Long-term and short-term memory network based movie comment sentiment analysis

This paper proposes an emotional analysis method of movie reviews based on Long-term and Short-term Memory(LSTM) Network model. Emotional analysis is widely used in movie recommendation system, which can recommend and judge movies by understanding the audiences emotional response to movies. However, due to the characteristics of movie text and the complexity of emotional expression, traditional methods such as machine learning have limitations and shortcomings in emotional analysis. However, the LSTM models better memory is utilized by the method proposed in this paper and the ability to capture the long-term correlation in movie texts, which obviously improves the accuracy and reliability of emotional analysis, and demonstrates the advantages of the LSTM model in emotional analysis compared to the traditional model. Future research can further explore other deep learning models and algorithms, so as to make emotional analysis more accurate and provide users with reliable movie recommendation information.

An Empirical Study on Tourism and Economic Growth in Greece: An Autoregressive Distributed Lag Boundary Test Approach

The objective of this research is to analyze the impact of the tourism industry on the economic growth of Greece. The study employs empirical analysis and time series econometric techniques to evaluate the Tourism-Driven Growth hypothesis. Information spanning from 1995 to 2022 about the growth of tourism (TR), expenditure on tourism (TE), average expenditure on tourism per capita (PCTE), and economic growth (GDP) was utilized. Initially, the authors examined the interconnections among these variables using the Autoregressive Distributed Lag (ARDL) Bounds Test. After identifying a statistically significant cointegration relationship, the study proceeded to estimate the long-term and short-term coefficients associated with these variables. Based on the results, it appears that there is a long-term correlation between economic growth and tourism, indicating that international tourism can have a positive impact on economic expansion.

The Age of the Smartphone: An Analysis of Social Predictors of Children’s Age of Access and Potential Consequences Over Time

Smartphones are the principal instrument for internet access among adolescents and pre-adolescents in many industrialized countries. However, research on the long-term correlates of age at first access to these devices concerning life outcomes is scarce. This study contributes to the literature by collecting data from 3,247 Italian students in grade 10. Through OLS and logistic regressions, we investigate socio-demographics’ role in predicting the age of first smartphone access and the associations between the age of access and selected life outcomes. The moderating effect of socio-demographics on such relationships is also investigated through multiple-interaction models. Results suggest that females and students from less-educated families are more likely to receive smartphones earlier. Early smartphone access is negatively associated with adolescents’ well-functioning. Finally, deferring access reduces the gender gap in language proficiency, digital skills and life satisfaction.

Degradation state estimation for insulated gate bipolar transistor based on multi-scale fusion learning

Insulated gate bipolar transistor (IGBT) modules are widely used in power electronic systems. Estimating the degradation state of IGBT modules is especially crucial when operating under different conditions, as they are highly likely to experience aging or faults, resulting in system failure. However, a fixed time window is generally used, which is not easy to fully learn the correlation between data. To address the issues, this paper proposes a multi-scale fusion learning method for estimating the degradation state of IGBT. The method can effectively mine the short-term and long-term correlations of IGBT aging data, which helps to implement the accurate estimation based on IGBT aging data. To validate the effectiveness of the proposed method, the accelerated aging test data provided by the NASA Predictive Center of Excellence (PCoE) is utilized. The results show that the method can accurately estimate the degradation state of IGBTs.

Macroeconomic determinants of the long-term correlation between stock and exchange rate markets in China: A DCC-MIDAS-X approach considering structural breaks

Owing to the liberalisation of financial markets, the impact of international capital flows on the Chinese stock market has become substantial. This study investigates the effects of economic policy uncertainty (EPU), geopolitical risk (GPR), consumer sentiment (CCI), macroeconomic fundamentals (MECI), and money supply (M2) on the correlations between the stock and exchange rate markets. The negative correlation between these two markets has become more pronounced in recent years. Moreover, EPU, GPR, CCI, and MECI negatively impact long-term stock-exchange rate correlations, while M2 has a positive impact. Portfolios of stock-exchange rates effectively reduce risk, especially when considering structural breaks.

Graph autoencoder with mirror temporal convolutional networks for traffic anomaly detection

Traffic time series anomaly detection has been intensively studied for years because of its potential applications in intelligent transportation. However, classical traffic anomaly detection methods often overlook the evolving dynamic associations between road network nodes, which leads to challenges in capturing the long-term temporal correlations, spatial characteristics, and abnormal node behaviors in datasets with high periodicity and trends, such as morning peak travel periods. In this paper, we propose a mirror temporal graph autoencoder (MTGAE) framework to explore anomalies and capture unseen nodes and the spatiotemporal correlation between nodes in the traffic network. Specifically, we propose the mirror temporal convolutional module to enhance feature extraction capabilities and capture hidden node-to-node features in the traffic network. Morever, we propose the graph convolutional gate recurrent unit cell (GCGRU CELL) module. This module uses Gaussian kernel functions to map data into a high-dimensional space, and enables the identification of anomalous information and potential anomalies within the complex interdependencies of the traffic network, based on prior knowledge and input data. We compared our work with several other advanced deep-learning anomaly detection models. Experimental results on the NYC dataset illustrate that our model works best compared to other models for traffic anomaly detection.

Deep causal speech enhancement and recognition using efficient long-short term memory Recurrent Neural Network.

Long short-term memory (LSTM) has been effectively used to represent sequential data in recent years. However, LSTM still struggles with capturing the long-term temporal dependencies. In this paper, we propose an hourglass-shaped LSTM that is able to capture long-term temporal correlations by reducing the feature resolutions without data loss. We have used skip connections in non-adjacent layers to avoid gradient decay. In addition, an attention process is incorporated into skip connections to emphasize the essential spectral features and spectral regions. The proposed LSTM model is applied to speech enhancement and recognition applications. The proposed LSTM model uses no future information, resulting in a causal system suitable for real-time processing. The combined spectral feature sets are used to train the LSTM model for improved performance. Using the proposed model, the ideal ratio mask (IRM) is estimated as a training objective. The experimental evaluations using short-time objective intelligibility (STOI) and perceptual evaluation of speech quality (PESQ) have demonstrated that the proposed model with robust feature representation obtained higher speech intelligibility and perceptual quality. With the TIMIT, LibriSpeech, and VoiceBank datasets, the proposed model improved STOI by 16.21%, 16.41%, and 18.33% over noisy speech, whereas PESQ is improved by 31.1%, 32.9%, and 32%. In seen and unseen noisy situations, the proposed model outperformed existing deep neural networks (DNNs), including baseline LSTM, feedforward neural network (FDNN), convolutional neural network (CNN), and generative adversarial network (GAN). With the Kaldi toolkit for automated speech recognition (ASR), the proposed model significantly reduced the word error rates (WERs) and reached an average WER of 15.13% in noisy backgrounds.

Fuelling the future : A comparative exploration of energy prices in India and the US

India’s crude oil consumption is expected to continue to increase, driven by economic growth and rising demand for transportation fuels. This will further boost India’s dependence on imported oil. The U.S. petroleum industry has several strengths, including its large resource base, technological expertise, and its access to capital. This paper analyses the association between economic conditions and the price of Crude Oil in a cross-country setting. The study uses a secondary data series of two prominent nations throughout 2013-2023. The analysis revealed a significant long-term price correlation between natural gas and crude oil, indicating a relationship between the prices of these energy commodities. The ramifications of the study’s conclusions hold significant relevance for investors and politicians. Investors should be aware that the cost of crude oil will likely be higher during economic uncertainty and inflation. Policymakers should also be mindful of Crude Oil’s role in the financial system and take steps to mitigate any potential risks.

Exploring the Long-Term Relationship between Foreign Direct Investment and GDP: Empirical Insights from Bangladesh

The study's purpose is to examine the long-term correlation between Foreign Direct Investment (FDI) and Gross Domestic Product (GDP) in Bangladesh. The article is mostly based on secondary data collected from Bangladesh Bank’s website covering for the period 1980-2021. Data were processed with the assistance of the Statistical Package of Social Science (SPSS IBM 23 Version). Correlation and regression analyses were used to examine the data in the study. The results suggest that there is a high degree of positive correlation (0.976) between FDI and GDP, along with the implication that FDI has a significant positive impact on GDP in the long run within the Bangladesh economy. The results also suggest that if FDI changes by 1 unit, then the GDP will positively change by 0.92381 at the level of 5% significant statistically. This finding is reinforced by the regression analysis, which indicates a statistically significant relationship between FDI and GDP growth. So, it is understandable that there is a well-built impact of FDI on Gross Domestic Product (GDP) in the long-run in the Bangladeshi economy. These results are further supported by regression analysis, underscoring a statistically significant relationship between FDI and GDP growth. However, it is important to acknowledge certain limitations, such as data availability constraints and potential confounding variables not addressed in the analysis. Future research endeavors could address these limitations and explore additional factors influencing the FDI-GDP relationship in Bangladesh. Overall, the study highlights the substantial impact of FDI on Gross Domestic Product in the long run and discusses policy recommendations to attract FDI inflows to Bangladesh.

An APT Attack Detection Method of a New-type Power System Based on STSA-transformer

Background: Complex structures such as a high proportion of power electronic equipment has brought new challenges to the safe and stable operation of new-type power system, increasing the possibility of the system being attacked, especially the more complex Advanced Persistent Threat (APT). This kind of attack has a long duration and strong concealment. Objective: Traditional detection methods target a relatively single attack mode, and the time span of APT processed is relatively short. None of them can effectively capture the long-term correlation in the attack, and the detection rate is low. These methods can’t meet the safety requirements of the new-type power system. In order to solve this problem, this paper proposes an improved transformer model called STSA-transformer algorithm, and applies it to the detection of APT in new-type power systems. Methods: In the STSA-transformer model, the network traffic collected from the power system is first converted into a sequence of feature vectors, and the location information and local feature of the sequence, is extracted by combining position encoding with convolutional embedding operations, and then global characteristics of attack sequences is captured using the multi-head selfattention mechanism of the transformer encoder, the higher-frequency features of the attention are extracted through the self-learning threshold operation, combined with the PowerNorm algorithm to standardize the samples, and finally classify the network traffic of the APT. Results: After multiple rounds of training on the model, the expected effect can be achieved and applied to the APT detection of a new-type power system. Conclusion: The experimental results show that the proposed STSA-transformer algorithm has better detection accuracy and lower detection false-alarm rate than traditional deep learning algorithms and machine learning algorithms.

QUANTIFYING THE COVID-19 SHOCK IN CRYPTOCURRENCIES

This paper sheds light on the changes suffered in cryptocurrencies due to the COVID-19 shock through a nonlinear cross-correlations and similarity perspective. We have collected daily price and volume data for the seven largest cryptocurrencies considering trade volume and market capitalization. For both attributes (price and volume), we calculate their volatility and compute the Multifractal Detrended Cross-Correlations (MF-DCCA) to estimate the complexity parameters that describe the degree of multifractality of the underlying process. We detect (before and during COVID-19) a standard multifractal behavior for these volatility time series pairs and an overall persistent long-term correlation. However, multifractality for price volatility time series pairs displays more persistent behavior than the volume volatility time series pairs. From a financial perspective, it reveals that the volatility time series pairs for the price are marked by an increase in the nonlinear cross-correlations excluding the pair Bitcoin versus Dogecoin [Formula: see text]. At the same time, all volatility time series pairs considering the volume attribute are marked by a decrease in the nonlinear cross-correlations. The K-means technique indicates that these volatility time series for the price attribute were resilient to the shock of COVID-19. While for these volatility time series for the volume attribute, we find that the COVID-19 shock drove changes in cryptocurrency groups.

CTFNet: Long-Sequence Time-Series Forecasting Based on Convolution and Time-Frequency Analysis.

Although current time-series forecasting methods have significantly improved the state-of-the-art (SOTA) results for long-sequence time-series forecasting (LSTF), they still have difficulty in capturing and extracting the features and dependencies of long-term sequences and suffer from information utilization bottlenecks and high-computational complexity. To address these issues, a lightweight single-hidden layer feedforward neural network (SLFN) combining convolution mapping and time-frequency decomposition called CTFNet is proposed with three distinctive characteristics. First, time-domain (TD) feature mining-in this article, a method for extracting the long-term correlation of horizontal TD features based on matrix factorization is proposed, which can effectively capture the interdependence among different sample points of a long time series. Second, multitask frequency-domain (FD) feature mining-this can effectively extract different frequency feature information of time-series data from the FD and minimize the loss of data features. Integrating multiscale dilated convolutions, simultaneously focusing on both global and local context feature dependencies at the sequence level, and mining the long-term dependencies of the multiscale frequency information and the spatial dependencies among the different scale frequency information, break the bottleneck of data utilization, and ensure the integrity of feature extraction. Third, highly efficient-the CTFNet model has a short training time and fast inference speed. Our empirical studies with nine benchmark datasets show that compared with state-of-the-art methods, CTFNet can reduce prediction error by 64.7% and 53.7% for multivariate and univariate time series, respectively.

The Impact of Institutional Performance and Environmental Sustainability on Foreign Direct Investment in Pakistan

In the context of Pakistan, this study aims to investigate how environmental sustainability, institutional performance, financial inclusion, and economic growth interact with and affect foreign direct investment .to determine if there are substantial causal linkages or correlations between these factors, as well as how these independent variables affect FDI inflows and outflows in Pakistan. The study also intends to offer insights on how individuals and governments may support FDI in the Pakistani setting. The research uses time-series data collected from 1997 to 2022 from the (ICRG) and (WDI) .It uses NARDL modelling to analyses the dynamic relationships between FDI, environmental sustainability, institutional performance, financial inclusion, and economic growth in the context of Pakistan. Unit root tests are used to evaluate each factor's stationarity. It seeks to clarify the short- and long-term correlations between these factors by doing this. The NARDL approach reveals substantial discrepancies in the connection among independent variables and foreign direct investment. The study finds connections between FDI and the selected independent variables that are both short- and long-term with positive and negative shocks. Foreign direct investment (FDI) benefits from institutional performance and economic development in both the short and long run, although it is less immediately influenced by financial inclusion and environmental sustainability. This highlights not just the need for more research to explore the collateral advantages of sustainability and financial inclusion, but also the necessity of supporting institutions and economic growth to attract foreign direct investment. To attract more foreign direct investment, Pakistan may really promote measures that support economic growth and institutional effectiveness. This adds to the body of scholarly research and offers useful advice to stakeholders and policymakers who want to maximize FDI inflows while fostering inclusive and sustainable economic growth in Pakistan.

Data-driven and knowledge-guided denoising diffusion model for flood forecasting

Data-driven models have been successfully applied in hydrological fields such as flood forecasting. However, limitations to the solutions to scientific problems still exist in this field: data collection is time-consuming and expensive, the quality of the collected data cannot be ensured, and noise or outliers may exist in the dataset, resulting in incorrect results. Moreover, data-driven models are trained only from available datasets and do not involve scientific principles or laws during the model-training process. This may lead to the prediction of specific scientific problems that do not conform to physical laws. Therefore, we propose a data-driven and knowledge-guided denoising diffusion (DK-Diffusion) model. First, for the data preprocessing stage, a coupled heterogeneous mapping tensor decomposition complementary algorithm is proposed that integrates the spatial information of a watershed, optimizes the initialization conditions of the model, reduces the potential correlation loss of data caused by tensor decomposition, and better optimizes the initial conditions of the model. We introduced an attention mechanism into the denoising diffusion probabilistic model (DDPM) to better capture medium and long-term correlations during flood processes. Most importantly, under the guidance of flood physics theory, we designed the loss function of the proposed model to ensure that the output prediction results were more consistent with the laws of flood physics. This is an innovative improvement with greater practical engineering value because it optimizes the boundary conditions of the model, giving it better generalization ability and reducing its dependence on data. Through comparative experiments on datasets from the Qijiang and Tunxi basins in China, compared with the popular flood forecasting model AGCLSTM, the root mean square error (RMSE) was reduced by 20.3–27.7%, and the mean absolute percentage error (MAPE) was reduced by 4.2–4.3%. Compared with the conditional score-based diffusion models for probabilistic time series imputation (CSDI), the average RMSE and mean sum of continuous ranked probability score CRPSsum were reduced by 6.3–10.6% and 6.1–6.2%, respectively.

Adaptive Spatio-Temporal Graph Learning for Bus Station Profiling

Understanding and managing public transportation systems require capturing complex spatio-temporal correlations within datasets. Existing studies often use predefined graphs in graph learning frameworks, neglecting shifted spatial and long-term temporal correlations, which are crucial in practical applications. To address these problems, we propose a novel bus station profiling framework to automatically infer the spatio-temporal correlations and capture the shifted spatial and long-term temporal correlations in the public transportation dataset. The proposed framework adopts and advances the graph learning structure through the following innovative ideas: (1) Designing an adaptive graph learning mechanism to capture the interactions between spatio-temporal correlations rather than relying on pre-defined graphs; (2) Modeling shifted correlation in shifted spatial graphs to learn fine-grained spatio-temporal features; (3) Employing self-attention mechanism to learn the long-term temporal correlations preserved in public transportation data. We conduct extensive experiments on three real-world datasets and exploit the learned profiles of stations for the station passenger flow prediction task. Experimental results demonstrate that the proposed framework outperforms all baselines under different settings and can produce meaningful bus station profiles.

Assessment of Food Security, Food Independence and Community Welfare in Food Insecure Areas

This study aims to analyze the factors that affect food security and its impact on the welfare and economic growth of Javanese society. This research method is correlation analysis with primary sources in the form of books and secondary sources in the form of articles and internet pages. This research is the result of a long-term correlation analysis as an effort to improve economic stability has an attachment between variables that affect the functions of Food Security, Food Independence, and Welfare which affect each other not enough with just one element standing alone. Food Security has a two tailed Pearson correlation that shows strong and positive results on food self-sufficiency and welfare. Food security, food self-sufficiency and welfare are very complex. The success of the program is also the shared responsibility of all components of the Nation and the Republic of Indonesia.

Study on the spatial and temporal correlation and allometric growth mechanism between population aging and carbon emissions in China.

Population aging and carbon emissions are critical issues for China's development. As an enormous complex system, the population and the carbon emission development process have non-negligible differences in time, space, and speed. Therefore, this paper first demonstrates the spatial and temporal correlation between population aging and carbon emissions from 1995 to 2020, then uses the allometric growth analysis model to make a cross-sectional temporal comparison and a vertical spatial comparison of the relationship and development rate of the two, and finally uses the ridge regression model to determine the forces and interaction mechanisms of the factors influencing the relationship between population aging and carbon emissions at allometric rates. The results show that (1) China has a long-term positive temporal correlation effect relationship between population aging and carbon emissions from 1995 to 2020, and the overall correlation is high. The spatial correlation intensity between population aging and carbon emissions varies significantly across Chinese provinces, with a general spatial distribution trend of high in the south, low in the north, and prominent in the center. (2) China's population aging and carbon emissions mainly show a negative allometric growth type of relationship, i.e., a strong trend of population aging expansion and a strengthening trend of carbon emission system shrinking. The number of provinces with negative allometric growth is gradually increasing, mainly in North, East, Central, and Southwest China. (3) From 1995-2010 period to the 2011-2020 period, the influence of the factors of the population, production, and economic dimensions on the population aging index and the carbon emission allometric scalar index gradually weakened, and the influence of the consumption and technology dimensions increased significantly. The factors on the population and consumption side of the dimension mainly contribute to the expansion of carbon emissions and drive positive allometric growth. The production side, the economic structure, and technology dimension factors drive negative allometric growth. The paper fully explores the bidirectional correlation, differential development trend, and interaction mechanism between the two systems of population and carbon emissions and effectively compensates for the lack of research content in terms of elemental correlation, spatial and temporal connection, and speed synergy.

Community Investments and Diet-Related Outcomes: A Longitudinal Study of Residents of Two Urban Neighborhoods

Investments in historically oppressed neighborhoods through food retail, housing, and commercial development are hypothesized to improve residents' health, nutrition, and perceptions of their neighborhood as a place to live. Although place-based development (e.g., housing, retail, business assistance) is happening in many communities, there is little evidence of the long-term correlates of multiple investments such as health and nutrition among residents. A quasi-experimental longitudinal study was conducted using a cohort of randomly sampled households in two low-income, predominantly African American neighborhoods in Pittsburgh, PA, with surveys assessing residents' food insecurity, perception of their neighborhood as a place to live, perception of access to healthy foods, and dietary outcomes in 2011 and seven years later (2018), with an interim assessment in 2014. Analyses conducted in 2022 compared changes among residents of one neighborhood which had 2.6 times the investments over a 7-year period with changes among residents of a socio-demographically similar neighborhood that received fewer investments. It was found that residents in the neighborhood receiving substantial investments demonstrated statistically significant improvements in neighborhood satisfaction (12.6% improvement compared with a 2.2% decrease) and perceived access to healthy food (52% improvement compared with 18.2% improvement), and marginally significant change in food security (14% compared with 4.8% improvement) compared with residents in the neighborhood receiving fewer investments. Multiple place-based investments in neighborhoods can potentially induce positive change for residents in health and nutrition outcomes.