Nature Of Time Series Research Articles

A model-free-based control chart for batch process using U-statistics

Batch processes are known to generate time series of successive measurements of many process variables in each run and the main challenge is to capture and accommodate the variability in the batch domain (batch-to-batch variability) and in the time domain (data with serial correlation). Classical approaches are focused on the first goal by applying multivariate techniques in the columns of a data matrix, in which each row has the data for the entire batch run. Those approaches are grounded in the traditional Multiway Principal Component Analysis (MPCA). Recent approaches are focused on the balance between the two sources, taking into account the time series nature of the data mainly by using time series models and neural network methods to directly accommodate the variability in the time domain. This paper proposes a model-free approach based on U-statistics theory. Through this theory, we can build a group of control charts (named V charts) capable of monitoring batch processes considering both sources of variability. In general, the new batch under monitoring is evaluated according to a group of reference in-control samples, based on the distance measures from those batches. The proposed approach is really flexible since we avoid model identification, parameter estimates issues and it is able to accommodate the within and between batch variability. Additionally, our method is suitable to deal with a wide range of batch data features, since under weak conditions, the distribution of the V statistics remains well known. It makes the V-based charts able to be used in different kinds of processes. We are focused here on controlling changes caused by data drifts (including trends) and by data serial correlation (data dynamics). Through the simulated and real batch process, we show that the V charts work well, including a scenario with very few reference batches available.

The Effect of Foreign Direct Investment on Improving the Life Expectancy Index in Iran

Background: The health care sector is a key sector of every country. In the process of economic development, this sector is considered infrastructure since economic development directly affects the productivity of the labor force and, as a result, the increase in production and economic growth. Therefore, the quality of the health care system may be an important factor in attracting foreign direct investment (FDI), accompanied by infrastructure such as labor, education, and research and development costs. As the level of health care increases in a society, the average life expectancy increases as well. Therefore, the life expectancy index (LEI) is an indicator for measuring the progress or backwardness of countries, and thus, it is necessary to examine the factors affecting it.
 Objectives: This study examined the effect of FDI on the LEI in Iran. Methods: The autoregressive distributed lags (ARDL) method was applied from 1981 to 2020, given the time series nature of the data and to test the research hypotheses. The method shows a capacity to be applied to I(0) and I(1) data simultaneously. the following regression model estimation was utilized to investigate the effect of FDI on LEI in Iran: HIt = c0 + c1 gross domestic product (GDP)t + c2 FDIt + c3 literacy rate (LR)t + c4 human immunodeficiency virus (HIV)t + c5 Urbent + c6 real exchange rate (EXR)t + c7 inflation rate (INF)t + c8 interest rate (IR)t + et.
 Results: The outcomes indicated that FDI has favorable and substantial impacts on the health expenditure variable in Iran over short and extended periods. As a result, it can be anticipated that investing in the health care industry should influence society’s economic advancement. Consequently, FDI becomes a significant macroeconomic factor in health care, acting as an extra investment source. With proper strategies, it holds the potential to serve as a crucial tool for the country’s progress, especially in enhancing the health care sector. The results demonstrated that FDI has constructive and notable effects on enhancing the LEI in Iran, both in the immediate and prolonged durations.
 Conclusions: Generally, the effects of FDI on health care can be described by direct and indirect channels. From the direct channel, FDI may affect health care outputs directly and indirectly, e.g., by providing a wider range of medical supplies and services (such as drugs and medical equipment), the expansion of FDI in the health care sector, and the presence of foreign firms producing medical or health care products that makes the supplies and services available at lower prices and consequently improves health care. From the indirect channel, FDI may influence the health care status of individuals in society indirectly by affecting other variables. Overall, based on the findings, life expectancy is likely to be improved in Iran by attracting more foreign investments in the health care sector.

Dividend Policy and Firm Performance: A Study of Selected Banks in Nigeria

This study investigated how the dividend policy affected the performance of a few Nigerian banks. Secondary data from the annual statements of the selected banks for the years 2010 to 2021 were used in the study. The study was guided by relevant specific objectives and hypotheses which were to test for the long run effects of the dividend policy and consider the effects of other determinants of bank performance for the period under study. The various preliminary tests and diagnostic tests were conducted to confirm that the data which was time series in nature were fit for empirical uses to avoid spurious results. The study utilized panel regression with the aid of E-views statistical package. The panel regression results indicates that the coefficient of the variables DEPOSIT, ASSET and Dividend indicate positive signs and are significant statistically. The positive coefficient of the variable DEPOSIT indicates that a percentage increase in the banks deposit increases the banks’ performance by 15%. For the variables ASSET and Dividend, the results show that a percentage increase in the bank’s asset and dividends increases the banks’ performance by 002% and 02% respectively. The results therefore confirmed that dividend policy by banks can enhance the overall performance of selected quoted Nigerian banks. Keywords: Dividend policy, firm performance, banks, Nigeria

STTRE: A Spatio-Temporal Transformer with Relative Embeddings for multivariate time series forecasting

The prevalence of multivariate time series data across several disciplines fosters a demand and, subsequently, significant growth in the research and advancement of multivariate time series analysis. Drawing inspiration from a popular natural language processing model, the Transformer, we propose the Spatio-Temporal Transformer with Relative Embeddings (STTRE) to address multivariate time series forecasting. This work primarily focuses on developing a Transformer-based framework that can fully exploit the spatio-temporal nature of a multivariate time series by incorporating several of the Transformer’s key components, but with augmentations that allow them to excel in multivariate time series forecasting. Current Transformer-based models for multivariate time series often neglect the data’s spatial component(s) and utilize absolute position embeddings as their only means to detect the data’s temporal component(s), which we show is flawed for time series applications. The lack of emphasis on fully exploiting the spatio-temporality of the data can incur subpar results in terms of accuracy. We redesign relative position representations, which we rename to relative embeddings, to unveil a new method for detecting latent spatial, temporal, and spatio-temporal dependencies more effectively than previous Transformer-based models. We couple these relative embeddings with a restructuring of the Transformer’s primary sequence learning mechanism, multi-head attention, in a way that allows for full utilization of relative embeddings, thus achieving up to a 24% improvement in accuracy over other state-of-the-art multivariate time series models on a comprehensive selection of publicly available multivariate time series forecasting datasets.

AN EMPIRICAL ANALYSIS OF TAX REVENUE BUOYANCY IN JAMMU AND KASHMIR

Abstract
 Taxation is a powerful instrument to finance the deficits of a country. Buoyancy of Tax is a degree of the sensitivity of tax incomes to economic growth. tax is said to be buoyant when revenues increase by more than one percent for a one percent rise in outputor national income (Audi, M. et. al. 2021). This paper is an attempt to examine the buoyancy of Tax revenue of J&K. The data is of time series nature covering the period from 2000- 01 to 2019-20 entirely relies on the secondary sources. The Augmented Dickey-fuller test has been used to check the unit root of the data. Ordinary least square method of regression has been used to estimate the buoyancy of tax revenue in Jammu and Kashmir. The results show that the buoyancy of all the major taxes is less than 1 except for other taxes(other taxes include many taxes).The implication is that the tax system did yield a less than 1 percent change resulting from economic activity, for every 1 percent change in GDP.

Analysis of aerosol optical depth and relation to covariates during pre-monsoon season (2002–2019) over Pakistan using ARIMAX model and cross-wavelet analysis

The pre-monsoon season heavily influences the precipitation amount in Pakistan. When hydrometeorological parameters interact with aerosols from multiple sources, a radiative climatic response is observed. In this study, aerosol optical depth (AOD) space-time dynamics were analyzed in relation to meteorological factors and surface parameters during the pre-monsoon season in the years 2002–2019 over Pakistan. Level-3 (L3) monthly datasets from Moderate Resolution Imaging Spectroradiometer (MODIS) and Multi-Angle Imaging Spectroradiometer (MISR) were used. Tropical Rainfall Measuring Mission (TRMM) derived monthly precipitation, Atmospheric Infrared Sounder (AIRS) derived air temperature, after moist relative humidity (RH) from Modern-Era Retrospective analysis for Research and Applications, Version-2 (MERRA-2), near-surface wind speed, and soil moisture data derived from Global Land Data Assimilation System (GLDAS) were also used on a monthly time scale. For AOD trend analysis, Mann-Kendall (MK) trend test was applied. Moreover, Autoregressive Integrated Moving Average with Explanatory variable (ARIMAX) technique was applied to observe the actual and predicted AOD trend, as well as test the multicollinearity of AOD with covariates. The periodicities of AOD were analyzed using continuous wavelet transformation (CWT) and the cross relationships of AOD with prevailing covariates on a time-frequency scale were analyzed by wavelet coherence analysis. A high variation of aerosols was observed in the spatiotemporal domain. The MK test showed a decreasing trend in AOD which was most significant in Baluchistan and Punjab, and the overall trend differs between MODIS and MISR datasets. ARIMAX model shows the correlation of AOD with varying meteorological and soil parameters. Wavelet analysis provides the abundance of periodicities in the 2–8 months periodic cycles. The coherency nature of the AOD time series along with other covariates manifests leading and lagging effects in the periodicities. Through this, a notable difference was concluded in space-time patterns between MODIS and MISR datasets. These findings may prove useful for short-term and long-term studies including oscillating features of AOD and covariates.

Attention-Based Deep Learning Modelling for Intrusion Detection

Cyber-attacks are becoming increasingly sophisticated, posing more significant challenges to traditional intrusion detection methods. The inability to prevent intrusions could compromise the credibility of security services, thereby putting data confidentiality, integrity, and availability at risk. In response to this problem, research has been conducted to apply deep learning (DL) models to intrusion detection, leveraging the new era of AI and the proven efficiency of DL in many fields. This study proposes a new intrusion detection system (IDS) based on DL, utilizing attention-based long short-term memory (AT-LSTM) and attention-based bidirectional LSTM (AT-BiLSTM) models. The time-series nature of network traffic data, which changes continuously over time, makes LSTM and BiLSTM particularly effective in handling intrusion detection. These models can capture long-term dependencies in the sequence of events, learn the patterns of normal network behaviour, and detect deviations from this behaviour that may indicate an intrusion. Also, the attention mechanism in the proposed models lets them make predictions based on the most important parts of the network traffic data. This is important for finding intrusions because network traffic data can have many different features, not all of which are important for finding an attack. The attention mechanism lets the models learn which features are most important for making accurate predictions, which improves their performance and efficiency. The UNSW-NB15 benchmark dataset is used in the study to measure and compare the effectiveness and reliability of the proposed system. This dataset contains normal and attack traffic data with a significant class imbalance. To address this issue, the study employs the Synthetic Minority Over-sampling Technique (SMOTE) to balance the dataset, thus reducing the risk of overfitting to the majority class and improving the model's performance in detecting attacks. The performance evaluation results demonstrate that the proposed models achieved a detection rate of over 93%, indicating high precision in detecting intrusions. By harnessing the power of deep learning, these models can learn and adapt to new threats over time, thus ensuring data confidentiality, integrity, and availability in today's interconnected world.

Biodiversity time series are biased towards increasing species richness in changing environments

The discrepancy between global loss and local constant species richness has led to debates over data quality, systematic biases in monitoring programmes and the adequacy of species richness to capture changes in biodiversity. We show that, more fundamentally, null expectations of stable richness can be wrong, despite independent yet equal colonization and extinction. We analysed fish and bird time series and found an overall richness increase. This increase reflects a systematic bias towards an earlier detection of colonizations than extinctions. To understand how much this bias influences richness trends, we simulated time series using a neutral model controlling for equilibrium richness and temporal autocorrelation (that is, no trend expected). These simulated time series showed significant changes in richness, highlighting the effect of temporal autocorrelation on the expected baseline for species richness changes. The finite nature of time series, the long persistence of declining populations and the potential strong dispersal limitation probably lead to richness changes when changing conditions promote compositional turnover. Temporal analyses of richness should incorporate this bias by considering appropriate neutral baselines for richness changes. Absence of richness trends over time, as previously reported, can actually reflect a negative deviation from the positive biodiversity trend expected by default.

THE EFFECT OF MICROFINANCE BANKS ON THE ECONOMIC GROWTH: EVIDENCE FROM NIGERIA

<p>The study seeks to examine the impact of microfinance institutions on the economic growth of a country, thus using Nigeria as a case study. The review utilizes the various relapse examination given that the information is cross-sectional and time series in nature. Optional information on all business banks was extricated from the National Bank of Nigeria’s measurable release and yearly reports. The information utilized in this model are time series optional information for the period 1992 to 2019. The discoveries of the review show that microfinance credits emphatically affect the short-run monetary presentation in Nigeria. Microfinance credits improved utilization per capita in the short run with a great coefficient, albeit these bank credits don't fundamentally affect monetary development over the long haul. Microfinance speculation be that as it may, essentially affects monetary execution in Nigeria over the long haul. Albeit miniature money advances are significant in the development process in Nigeria, different measures, for example, helping farming creation and finding a way suitable way to upgrade per capita pay are similarly significant in supporting the Nigerian monetary development. That's what we suggest, microfinance foundations ought to credit to further develop utilization in the short run, while the long run objective ought to be to further develop speculation and other capital aggregation.</p><p> </p><p><strong>JEL</strong>: D01, D04, G21</p><p> </p><p><strong> Article visualizations:</strong></p><p><img src="/-counters-/soc/0001/a.php" alt="Hit counter" /></p>

DIGITAL FINANCE AND FINANCIAL INCLUSION IN NIGERIA: ARE THEY SIAMESE TWINS?

The link between Financial inclusion (FIN ) and digital finance(DF) is the primary subject of this research. Digital finance and inclusive financial though have been adjudged to be related; there are quite a plethora of issues that are still unclear concerning the nature and type of their relationship. Why are they related and at what point are they related? Short or long-run relationship? To clarify these issues, the study applied the Auto regressive Distributed Lag (ARDL) long-run cointegrating technique to explicate our econometric models. Our data is time series in nature and is for Nigeria. The study established that, internet-powered electronic device users, a proxy for digital finance has a positive long-run relationship with FIN while Automated Teller machine (ATM) ownership and usage has a negative long-run relationship with FIN. The study recommended that internet service providers (ISPs) be integrated into the Nigerian FIN paradigm to enhance the accuracy and viability of financial service provision, leading to a decline in the unbanked population.
 Keywords: Digital finance, FIN, financial services, ARDL Bounds Test, Nigeria.

Fiscal Policy, Private Investment and Manufacturing Sector Performance: Empirical Evidence from Nigeria with ARDL Technique

One of the greatest challenges of the Sub-Saharan African economies today is how to develop the appropriate model that will propel development in the continent. The manufacturing sector instances in China, Malaysia, Taiwan, Korea and Japan are clear models to adopt for Nigeria. The question now is, what has Nigeria been doing to develop the manufacturing sector? To this end, this study set out to examine the effect of fiscal policy and private investment on the performance of the manufacturing sector in Nigeria from 1981 to 2021. The study is time series in nature and made use of the autoregressive distributed lag (ARDL) modelling technique. It was found that domestic private investment and aggregate government spending exert a significant positive influence on the manufacturing sector performance in the long run, while the influence from tax and government revenue was significantly negative. It was then recommended that government should improve on the factors that attract more FDI inflows into the country; boosts requisite educational training skills and financial market depth to drive the influence of domestic private investment on the manufacturing sector; increase government spending on the manufacturing sector development; and that tax revenues be judiciously utilized in the long-run so that they can positively impact the manufacturing sector.

Short-Term Power Load Forecasting Based on an EPT-VMD-TCN-TPA Model

Accurate short-term load forecasting is the key to ensuring smooth and efficient power system operation and power market dispatch planning. However, the nonlinear, non-stationary, and time series nature of load sequences makes load forecasting difficult. To address these problems, this paper proposes a short-term load forecasting method (EPT-VMD-TCN-TPA) based on the hybrid decomposition of load sequences, which combines ensemble patch transform (EPT), variational modal decomposition (VMD), a temporal convolutional network (TCN), and a temporal pattern attention mechanism (TPA). In which, the trend component (Tr(t)) and the residual fluctuation component (Re(t)) of the load series are extracted using EPT, and then the Re(t) component is decomposed into intrinsic modal function components (IMFs) of different frequencies using VMD. The Tr(t) and IMFs components of the fused meteorological data are predicted separately by the TCN-TPA prediction model, and finally, the prediction results of each component are reconstructed and superimposed to obtain the final predicted value of the load. In addition, experiments after reconstructing each IMF component according to the fuzzy entropy (FE) values are discussed in this paper. To evaluate the performance of the proposed method in this paper, we used datasets from two Areas of the 9th Mathematical Modeling Contest in China. The experimental results show that the predictive precision of the EPT-VMD-TCN-TPA model outperforms other comparative models. More specifically, the experimental results of the EPT-VMD-TCN-TPA method had a MAPE of 1.25% and 1.58% on Area 1 and Area 2 test sets, respectively.

RF-BiLSTM Neural Network Incorporating Attention Mechanism for Online Ride-Hailing Demand Forecasting

Accurately predicting online ride-hailing demand can help operators allocate vehicle resources on demand, avoid idle time, and improve traffic conditions. However, due to the randomness and complexity of online ride-hailing demand data, which are affected by many factors and mostly time-series in nature, it is difficult to forecast accurately and effectively based on traditional forecasting models. Therefore, this study proposes an online ride-hailing demand forecasting model based on the attention mechanism of a random forest (RF) combined with a symmetric bidirectional long short-term memory (BiLSTM) neural network (Att-RF-BiLSTM). The model optimizes the inputs and can use past and future data to forecast, improving the forecasting precision of online ride-hailing demand. The model utilizes a random forest to filter and optimize the input variables to reduce the neural network complexity, and then an attention mechanism was incorporated into the BiLSTM neural network to construct a demand forecasting model and validate it using actual Uber pickup data from New York City. Compared with other forecasting models (Att-XGBoost-BiLSTM, Att-BiLSTM, and pure LSTM), the results show that the proposed symmetrical Att-RF-BiLSTM online ride-hailing demand forecasting model has a higher forecasting precision and fitting degree, which indicates that the proposed model can be satisfactorily applied to the area of online ride-hailing demand.

A recursive multi-head graph attention residual network for high-speed train wheelset bearing fault diagnosis

Wheelset bearings are a core component of high-speed trains, and their fault diagnosis is the key to smooth operation. Deep learning is widely used in fault diagnosis due to its powerful classification ability. To explicitly fit the features of vibration signals and further explore the relationship between the signals, the graph attention network (GAT) is becoming a focus of research. Unlike traditional graph neural networks, GATs can focus on edges with stronger correlations with vertices, making the model more powerful when fitting graph samples in non-Euclidean space. However, existing GATs have two limitations. Firstly, most graph construction methods do not consider the characteristics of vibration signals, so the graph interpretation is not good. Secondly, the existing methods of graph attention coefficient cannot effectively reflect the importance of edges. To address these issues, a recursive multi-head graph attention residual network (RMHGARN) is proposed. In RMHGARN, vibration signals are transformed into recurrence graphs due to the recursive nature of nonlinear time series. A multi-kernel Gaussian symmetric graph attention mechanism is proposed to obtain the Hilbert spatial distribution between neighboring vertices. In addition, a graph encoding module is proposed to improve the feature representation of input samples. The effectiveness and superiority of RMHGARN under strong noise samples are verified using three wheelset bearing datasets.

Deep Ensemble-Based Approach Using Randomized Low-Rank Approximation for Sustainable Groundwater Level Prediction

Groundwater is the most abundant freshwater resource. Agriculture, industrialization, and domestic water supplies rely on it. The depletion of groundwater leads to drought. Topographic elevation, aquifer properties, and geomorphology influence groundwater quality. As the groundwater level data (GWL) are time series in nature, it is challenging to determine appropriate metrics and to evaluate groundwater levels accurately with less information loss. An effort has been made to forecast groundwater levels in India by developing a deep ensemble learning approach using a double-edge bi-directed long-short-term-memory (DEBi-LSTM) model approximated with a randomized low-ranked approximation algorithm (RLRA) and the variance inflation factor (VIF) to reduce information loss and to preserve data consistency. With minimal computation time, the model outperformed existing state-of-the-art models with 96.1% accuracy. To ensure sustainable groundwater development, the proposed work is discussed in terms of its managerial implications. By applying the model, we can identify safe, critical, and semi-critical groundwater levels in Indian states so that strategic plans can be developed.

Graph Deep Factors for Probabilistic Time-series Forecasting

Effective time-series forecasting methods are of significant importance to solve a broad spectrum of research problems. Deep probabilistic forecasting techniques have recently been proposed for modeling large collections of time-series. However, these techniques explicitly assume either complete independence (local model) or complete dependence (global model) between time-series in the collection. This corresponds to the two extreme cases where every time-series is disconnected from every other time-series in the collection or likewise, that every time-series is related to every other time-series resulting in a completely connected graph. In this work, we propose a deep hybrid probabilistic graph-based forecasting framework called Graph Deep Factors (GraphDF) that goes beyond these two extremes by allowing nodes and their time-series to be connected to others in an arbitrary fashion. GraphDF is a hybrid forecasting framework that consists of a relational global and relational local model. In particular, a relational global model learns complex non-linear time-series patterns globally using the structure of the graph to improve both forecasting accuracy and computational efficiency. Similarly, instead of modeling every time-series independently, a relational local model not only considers its individual time-series but also the time-series of nodes that are connected in the graph. The experiments demonstrate the effectiveness of the proposed deep hybrid graph-based forecasting model compared to the state-of-the-art methods in terms of its forecasting accuracy, runtime, and scalability. Our case study reveals that GraphDF can successfully generate cloud usage forecasts and opportunistically schedule workloads to increase cloud cluster utilization by 47.5% on average. Furthermore, we target addressing the common nature of many time-series forecasting applications where time-series are provided in a streaming version; however, most methods fail to leverage the newly incoming time-series values and result in worse performance over time. In this article, we propose an online incremental learning framework for probabilistic forecasting. The framework is theoretically proven to have lower time and space complexity. The framework can be universally applied to many other machine learning-based methods.

Two models for predicting stock prices in combination with LSTM

In addition to its practical and theoretical significance, stock forecasting has long been a hot research topic for scholars domestically and abroad. Stock data are time-series in nature, and neural networks have achieved relatively good performance in dealing with time series problems, among which long-short-term memory neural networks are well suited to dealing with such time-series data with long-term dependence. However, the stock market is an environment that changes with the external environment, with high stochasticity and complex intrinsic nonlinear relationships between different phenomena. Relying on a single method to identify the series directly cannot fully extract the complex information of the series changes, so the combined forecasting method is proposed. One idea is to combine empirical mode decomposition (EMD) with long-short-term memory (LSTM), and another idea is to incorporate LSTM or its variant GRU into generative adversarial networks (GAN/CGAN/WGAN). After an empirical study of Guanhao Bio's stock price and Apple's stock price, both methods present better prediction results.

The relationship between the structure of the board of directors and firm performance in family versus non-family firms

This paper investigates the impact of four characteristics of the board of directors (board size, board independence, leadership structure and board meetings) on firm performance (Tobin's q). Accordingly, four hypotheses have been developed. The analysis is based on data (objective variables) from 221 firms operating in three countries (Spain, Portugal and Italy) and differentiates between non-family-controlled businesses (NFCBs) and family-controlled businesses (FCBs). Considering the cross-sectional (three countries) and the time series (six years) nature of the data, we used a panel data estimation approach. Our findings show that, although corporate governance recommendations advocate larger, more independent and proactive boards, as well as structures that ensure the separation of the chairperson and CEO roles, these board features do not always result in more effective boards. Indeed, smaller, less independent and less active boards and dual leadership structures are tied to better performance by the FCBs in our sample as compared with the NFCBs.

Real-Time Personalized Physiologically Based Stress Detection for Hazardous Operations

When training for hazardous operations, real-time stress detection is an asset for optimizing task performance and reducing stress. Stress detection systems train a machine-learning model with physiological signals to classify stress levels of unseen data. Unfortunately, individual differences and the time-series nature of physiological signals limit the effectiveness of generalized models and hinder both post-hoc stress detection and real-time monitoring. This study evaluated a personalized stress detection system that selects a personalized subset of features for model training. The system was evaluated post-hoc for real-time deployment. Further, traditional classifiers were assessed for error caused by indirect approximations against a benchmark, optimal probability classifier (Approximate Bayes; ABayes). Healthy participants completed a task with three levels of stressors (low, medium, high), either a complex task in virtual reality (responding to spaceflight emergency fires, n=27) or a simple laboratory-based task (N-back, n=14). Heart rate, blood pressure, electrodermal activity, and respiration were assessed. Personalized features and different interval window sizes were compared. Classification performance was compared for ABayes, support vector machine, decision tree, and random forest. The results demonstrate that a personalized model with time series intervals can classify three stress levels with high accuracy better than a generalized model. However, cross-validation and holdout performance varied for traditional classifiers vs. ABayes, suggesting error from indirect approximations. The selected features changed with windows size and differed between tasks, but blood pressure contributed to the most prominent features. The capability to account for individual difference is a prominent advantage of personalized models, and will likely have a growing presence in future detection systems.

Development of wavelet-based Kalman Online Sequential Extreme Learning Machine optimized with Boruta-Random Forest for drought index forecasting

Drought is a stochastic and recurring hydrological natural hazard that occurs due to a shortage of precipitation over a period of time. Drought forecasting in water resources systems has an important role in reducing devastating ecological and social impacts. However, due to the fluctuating nature of drought indicator time series, their forecasting on a short time scale without advanced pre-processing is extremely challenging, and little research has been presented in this field. In this study, a new complementary machine learning (ML) approach, Kalman filter regression-based Online Sequential Extreme Learning Machine and (KOSELM) coupled with the Boundary Corrected Maximal Overlap Discrete Wavelet Transform (BC-MODWT-KOSELM), was implemented for forecasting one-month/three-month ahead Standardized Precipitation Evapotranspiration Index (i.e., SPEI 12 and SPEI 24). Here, the Kalman filter regression was employed to optimize the hyper-parameters of the Online Sequential Extreme Learning Machine (OSELM). Precipitation and potential evapotranspiration data from Bandar Abbas (warm semi-humid climate) and Rasmar (humid climate) synoptic stations, Iran were used in the analysis for a period of (1987–2019). In order to validate the BC-MODWT-KOSELM model, it was compared with two other well-known ML approaches, classical Extreme Learning Machine (ELM) and Generalized Regression Neural Network (GRNN) in both standalone and BC-MODWT-based complementary frameworks (i.e., BC-MODWT-ELM and BC-MODWT-GRNN). First, the original time series of the benchmark drought index was decomposed into the wavelet and scaling coefficients based on three mother wavelets and different decomposition levels. Afterward, the most significant lags were extracted using the Boruta-random forest (B-RF) feature selection. The BC-MODWT-KOSELM was identified as the superior forecasting model for SPEI 12 (t ＋ 1) with R = 0.9511 and RMSE = 0.3318; for SPEI 12 (t ＋ 3) with R = 0.8171 and RMSE = 0.6168; for SPEI 24 (t ＋ 1) with R = 0.9710 and RMSE = 0.2156; for SPEI 24 (t ＋ 3) with R = 0.9014 and RMSE = 0.3937 in the Bandar Abbas station. Besides, the BC-MODWT-KOSELM outperformed the comparative counterpart model for SPEI 12 (t ＋ 1 with R = 0.9401 and RMSE = 0.3371; SPEI 12 (t ＋ 3) with R = 0.8266 and RMSE = 0.5723; for SPEI 24 (t ＋ 1) with R = 0.9640 and RMSE = 0.2974; for SPEI 24 (t ＋ 3) with R = 0.9063 and RMSE = 0.4874 in the Ramsar station.