Series Data Research Articles

Theoretical Notions with Practical Applications Regarding the Statistical Calculation of Dental Bur Wear

This paper presents theoretical notions with practical applications for establishing the lifetime of a dental bur (durability and reliability) based on the wear of the active part and considering a series of statistical indicators. To justify applying theoretical notions in practice, a conical–cylindrical dental bur is studied experimentally to obtain a series of data of the work process necessary for statistical calculation. The parameter taken into consideration in this paper is mass lost through the wear of the dental bur active part, based on testing. This is useful for dental bur lifetime establishment and its extension or even optimization in operation. The loss of mass of the dental bur active part is analyzed in the work process using the results and experimental data obtained and validated by statistical–mathematical numerical calculation. The numerical calculation approximates the mass lost through wear at different rotation speeds and operation times, and based on a comparison with the experimentally determined ones, the lifetime was established. The results show that the dental bur works with high yield in the first 20 h of work, after which it should be replaced with a new one. Theoretically, the studied dental bur can work over 21 h. Practically, it can work up to 18 h without major risk of failure, but the lifetime can be extended up to 20 h, at which point the failure risk can reach 10% and it is recommended to replace the bur.

A Deep–Learning Network for Wheat Yield Prediction Combining Weather Forecasts and Remote Sensing Data

Accurately predicting winter wheat yield before harvest could greatly benefit decision-makers when making management decisions. In this study, we utilized weather forecast (WF) data combined with Sentinel-2 data to establish the deep-learning network and achieved an in-season county-scale wheat yield prediction in China’s main wheat-producing areas. We tested a combination of short-term WF data from the China Meteorological Administration to predict in-season yield at different forecast lengths. The results showed that explicitly incorporating WF data can improve the accuracy in crop yield predictions [Root Mean Square Error (RMSE) = 0.517 t/ha] compared to using only remote sensing data (RMSE = 0.624 t/ha). After comparing a series of WF data with different time series lengths, we found that adding 25 days of WF data can achieve the highest yield prediction accuracy. Specifically, the highest accuracy (RMSE = 0.496 t/ha) is achieved when predictions are made on Day of The Year (DOY) 215 (40 days before harvest). Our study established a deep-learning model which can be used for early yield prediction at the county level, and we have proved that weather forecast data can also be applied in data-driven deep-learning yield prediction tasks.

A century-long China homogenized daily surface air temperature dataset (CUG-CMA CHDT)

Daily meteorological observation data of the early period (pre-1950) were critically important for investigating the long-term trends and multi-decadal scale variability of extreme climate events. The high-resolution surface air temperature (SAT) data for time period before 1950 are lacking in China. We extended the SAT observations of China back to 1840 through developing a pre-1950 daily SAT dataset. The early-period daily SAT were manually corrected for the input and clerical errors, and then according to the length or coverage of time, the main series for each of the cities was determined. The observation time system of unknown sites was determined by the minimum difference method. After these operations, the data of all sites were unified into the same format. By using the ridge regressions established based on data from modern reference stations, the missing maximum temperature (Tmax) and minimum temperature (Tmin) were interpolated. The early-period data were combined with modern data to form the long-term daily SAT dataset of 1840–2020 in China. RHtest software was used to detect and adjust the inhomogeneities in the station data series. Finally, the century-long homogenized daily SAT dataset including 45 key city stations in China was obtained. Among the stations, there are 20 stations with observation record more than one hundred years. The length of temperature observation series of 17 stations is between 80 and 100 years. The series length of the remaining 7 stations is between 68 and 80 years. Finally, the angular distance weighting (ADW) method was used to interpolate the data into grid products, and the grid size is 2.5 ° × 2.5 °. The dataset was named CUG-CMA CHDT, which is applicable in monitoring, studies and assessments of regional extreme temperature change and variability in China.

Zoonotic outbreak risk prediction with long short-term memory models: a case study with schistosomiasis, echinococcosis, and leptospirosis

BackgroundZoonotic infections, characterized with huge pathogen diversity, wide affecting area and great society harm, have become a major global public health problem. Early and accurate prediction of their outbreaks is crucial for disease control. The aim of this study was to develop zoonotic diseases risk predictive models based on time-series incidence data and three zoonotic diseases in mainland China were employed as cases.MethodsThe incidence data for schistosomiasis, echinococcosis, and leptospirosis were downloaded from the Scientific Data Centre of the National Ministry of Health of China, and were processed by interpolation, dynamic curve reconstruction and time series decomposition. Data were decomposed into three distinct components: the trend component, the seasonal component, and the residual component. The trend component was used as input to construct the Long Short-Term Memory (LSTM) prediction model, while the seasonal component was used in the comparison of the periods and amplitudes. Finaly, the accuracy of the hybrid LSTM prediction model was comprehensive evaluated.ResultsThis study employed trend series of incidence numbers and incidence rates of three zoonotic diseases for modeling. The prediction results of the model showed that the predicted incidence number and incidence rate were very close to the real incidence data. Model evaluation revealed that the prediction error of the hybrid LSTM model was smaller than that of the single LSTM. Thus, these results demonstrate that using trending sequences as input sequences for the model leads to better-fitting predictive models.ConclusionsOur study successfully developed LSTM hybrid models for disease outbreak risk prediction using three zoonotic diseases as case studies. We demonstrate that the LSTM, when combined with time series decomposition, delivers more accurate results compared to conventional LSTM models using the raw data series. Disease outbreak trends can be predicted more accurately using hybrid models.

Population age structure dependency of the excess mortality P-score

BackgroundSince the outbreak of the COVID-19 pandemic, the excess mortality P-score has gained prominence as a measure of pandemic burden. The P-score indicates the percentage by which observed deaths deviate from expected deaths. As the P-score is regularly used to compare excess mortality between countries, questions arise regarding the age dependency of the measure. In this paper we present formal and empirical results on the population structure bias of the P-score with a special focus on cross-country comparisons during the COVID-19 pandemic in Europe.MethodsP-scores were calculated for European countries for 2021, 2022, and 2023 using data from the 2024 revision of the United Nations’ World Population Prospects and the HMDs Short Term Mortality Fluctuations data series. The expected deaths for 2021, 2022, and 2023 were estimated using a Lee–Carter forecast model assuming pre-pandemic conditions. P-score differences between countries were decomposed using a Kitagawa-type decomposition into excess-mortality and structural components. To investigate the sensitivity of P-score cross-country rankings to differences in population structure we calculated the rank-correlation between age-standardized and classical P-scores.ResultsThe P-score is an average of age-specific percent excess deaths weighted by the age-distribution of expected deaths. It can be shown that the effect of differences in the distribution of deaths only plays a marginal role in a European comparison. In most cases, the excess mortality effect is the dominant effect. P-score rankings among European countries during the COVID-19 pandemic are similar under both age-standardized and classical P-scores.ConclusionsAlthough the P-score formally depends on the age-distribution of expected deaths, this structural component only plays a minor role in a European comparison, as the distribution of deaths across the continent is similar. Thus, the P-score is suitable as a measure of excess mortality in a European comparison, as it mainly reflects the differences in excess mortality. However, this finding should not be extrapolated to global comparisons, where countries could have very different death distributions. In situations were P-score comparisons are biased age-standardization can be applied as a solution.

Predictive ARIMA Model with a Machine Learning (ML) Approach for COVID-19 Data in Pakistan

This study is based on the application of an ARIMA (p, d, q) based machine learning (ML) approach to evaluate the dynamics of COVID-19 pandemic. The focus is on estimating epidemic trends and performing diagnostic scrutiny with model fitting. The data including all four waves of the pandemic pertaining to Pakistan, covering all four provinces (Sindh, Punjab, Khyber Pakhtunkhwa, Balochistan, as well as Gilgit Baltistan, Azad Jammu Kashmir, and the capital city Islamabad, collected from February 26, 2020, to September 30, 2021, is analyzed. The ML algorithm is used to optimize the results of ADF, unit root test which ensures the minimum of ACF, and PACF graphs intention of the data series. The results employ the fitted ARIMA models (1, 1, 1) and (1, 1, 7) for the 1st to 4th waves, confirming daily infected cases across the entire dataset of Pakistan. The cumulative trained observations are from the 1st wave (February 26, 2020, to October 20, 2020), 2nd wave (October 21, 2020, to March 16, 2021), 3rd wave (March 17, 2021, to July 10, 2021), and 4th wave (July 11, 2021, to September 30, 2021), with a further 14-day forecast (from October 1 to October 14, 2021). The results show a strong correlation between the trained and predicted values, ranging from 0.8789 to 0.99236. To select predictive model parameters, the model that results in the minimum Bayesian Information Criterion (BIC) value and residuals from the datasets obtained after detaching the unnecessary errors and the 95^% CI for the forecasting error ( ) are calculated. These values would help to decide the best fitted predictive model.

Volume, increment, and aboveground biomass data series and biomass conversion and expansion factors for the main forest types of EU Member States

Key messageThis collection reports the standing stock volume, increment, aboveground biomass, and biomass conversion and expansion factors attributed to 222 forest types and 48 different management types, representative of 25 EU Member States. DOI: https://zenodo.org/records/11387301.

Continuity of long-term climate data series after the transition from manual to automatic weather station

In climatology, transitioning from conventional manual weather stations (MWS) to automatic weather stations (AWS) presents challenges in maintaining the homogeneity of long-term data series. At the Skalnaté Pleso Observatory (1778 m a.s.l., High Tatras), manual meteorological measurements have been conducted since 1943 using the same methods and devices at the same location. In 2014, the AWS Physicus was installed to record meteorological data parallelly with manual measurements. This study processed six years of parallel measurements (2017–2022) to derive corrections for AWS monthly atmospheric precipitation totals and maximum (Tmax), minimum (Tmin) and mean air temperatures (Tmean) measured in 2023. Two correction approaches were proposed: monthly regressions (MR) and cumulative distribution functions (CDF), including Generalized Extreme Value (GEV) distribution, normal distribution (GAUSS), and Gamma distribution (GAMMA). Analyses of monthly data revealed an underestimation of precipitation totals in AWS data, with a mean bias error (MBE) of −6.8 mm and a root mean squared error (RMSE) of 20.0 mm. The most suitable correction method was the MR, which decreased MBE to 1.5 mm and RMSE to 11.1 mm. AWS monthly air temperature data were overestimated by 0.1 °C, 0.3 °C, and 0.1 °C for Tmax, Tmin, and Tmean, respectively. For Tmax, correction using the MR and GEV methods reduced the MBE to 0.0 °C and achieved RMSE of 0.1 °C for both. After applying GAUSS, MBE and RMSE decreased to 0.1 °C. The most appropriate correction method for Tmin was the MR resulting in MBE of 0.0 °C and RMSE of 0.3 °C. For Tmean, the MR and GEV methods reduced the MBE to 0.0 °C and the RMSE to 0.1 °C for both methods. These results demonstrate that AWS monthly data, when corrected using the presented methods, have the potential to maintain the continuity of historical climate data series at Skalnaté Pleso.

A novel reservoir dispatching rules extraction framework based on hybrid embedding informer

Accurately capturing reservoir dispatching rules is crucial for mitigating carbon emissions and preventing flooding and inundation. However, the short operational duration of certain reservoirs poses challenges in acquiring sufficient data samples for reservoir rule simulation. Furthermore, the nonlinear and fluctuating characteristics of reservoir dispatching rules present obstacles to their precise prediction. To address these challenges, this study proposes a hybrid embedding-based Informer (Heformer) dispatching rules extraction framework. Initially, extensive data series are employed for modeling to acquire a comprehensive sample dataset. Subsequently, the Heformer model, based on the Informer architecture and integrating the Dimension-Segment-Wise (DSW) strategy and feature representation strategy, is introduced for capturing reservoir dispatching rules. To assess performance of the proposed model, it is compared against several recognized state-of-the-art models. Experimental results demonstrate that the proposed method effectively captures the operational characteristics of reservoirs, mitigates bias impact in reservoir prediction and operation, and provides a reference for extracting dispatching rules for reservoirs.

Effect of Meteorological Variables on Energy Demand in the Northeast and Southeast Regions of Brazil

Energy consumption demand has shown successive records during recent months, primarily associated with heat waves in almost all Brazilian states. The effects of climate change induced by global warming and the increasingly frequent occurrence of extreme events, mainly regarding temperature and precipitation, are associated with this increase in demand. In this sense, the impact of meteorological variables on load demand in some substations in the northeast and southeast of Brazil was analyzed, considering the historical series of energy injected into these substations. Fifteen substations were analyzed: three in the state of São Paulo, six in Bahia, three in Pernambuco, and three in Rio Grande do Norte. Initially, essential quality control was carried out on the energy injection data. The SAMeT data sets were used for the variable temperature, and Xavier was used for precipitation and relative humidity to obtain a homogeneous data series. Daily and monthly data were used for a detailed analysis of these variables in energy demand over the northeast and southeast regions of Brazil. Some regions were observed to be sensitive to the maximum temperature variable, while others were sensitive to the average temperature. On the other hand, few cases showed the highest correlation with the precipitation and relative humidity variables, with most cases being considered slight or close to zero. A more refined analysis was based on the type of consumers associated with each substation. These results showed that where consumption is more residential, the highest correlations were associated with maximum temperature in most stations in the northeast and average temperature in the southeast. In regions where consumption is primarily rural, the correlation observed with precipitation and relative humidity was the highest despite being negative. A more detailed analysis shows that rural production is associated with irrigation in these substations, which partly explains consumption, as when rainfall occurs, the demand for irrigation decreases, and thus energy consumption is reduced.

An automated Peak Group Analysis for vibrational spectra analysis

Peak Group Analysis (PGA) is a multivariate curve resolution technique that attempts to extract single pure component spectra from time series of spectral mixture data. It requires that the mixture spectra consist of relatively sharp peaks, as is typical in IR and Raman spectroscopy. PGA aims to construct from individual peaks the associated pure component spectra in the form of nonnegative linear combinations of the right singular vectors of the spectral data matrix.This work presents an automated PGA (autoPGA) that starts with upstream peak detection applied to time series of spectra, combining different window-based peak detection techniques with balanced peak acceptance criteria and peak grouping to deal with repeated detections. The next step is a single-spectrum-oriented PGA analysis. This is followed by a downstream correlation analysis to identify pure component spectra that occur multiple times. AutoPGA provides a complete pure component factorization of the matrix of measured data. The algorithm is applied to FT-IR data sets on various rhodium carbonyl complexes and from an equilibrium of iridium complexes.

An extension to Kendall’s Tau metric to evaluate dissimilarities between data series

Data analysis is performed to examine, interpret, and extract information from data series, and it includes applying various methods and techniques to understand patterns and compare data. An approach to compare data is to use rank metrics that help identify how distinct two data series are when compared to each other according to patterns, formats, criteria, and dimensions in both data series. Among these metrics, Kendall’s Tau metric stands out, as it is robust and inexpensive, widely used in analyzing sequences and genomes, to detect errors in flash memories, and to compare distributions and top-k ranked values. However, a challenge arises when comparing lists with different lengths or when lists do not share the same elements. This happens, for example, when lists are defined by top-k elements, commonly called k-list. In this case, there is no guarantee that two k-lists share the same set of elements. Traditional metrics like Kendall’s Tau are designed to quantify differences only between shared elements in lists. Recognizing this limitation, a possible solution is to apply the metric to the shared elements of the lists. Another solution, named the generalization of Kendall’s Tau, proposed by Fagin et al., considers all elements in two lists. However, this generalization of Kendall Tau is a semi-metric, as it does not satisfy the triangular inequality. To solve this problem, we propose the Extended Kendall Tau (EKT) metric that meets all the conditions of a metric and simultaneously considers the distinct elements of the compared lists. The proposed metric was evaluated by applying conventional Kendall’s Tau and the extended Kendall’s Tau over 40 text files divided into five different languages (eight files per language). We compared KT and EKT measures within the ”same language” and across ”other language” files for the two scenarios. The results revealed that both methods could accurately identify the differences between the groups of texts of the ”same language” and ”other language”. However, the numerical results show that EKT is able to more significantly highlight the difference between groups of texts of different languages.

Rapid identification and early warning of axial compressor stall based on multiscale CNN-SVM-FC model

Early prewarning of compressor stall and surge is crucial to avoid aircraft engine instability, yet it is challenging due to the complex and unstable flow field characterized by multiple modes and multiscale features. To enhance the multi-scale feature representation capability of Convolutional Neural Network-Support Vector Machine (CNN-SVM) algorithm, a novel classifier modelling method combined multiscale windows with CNN-SVM is introduced for stall prewarning in this paper, named Multiscale CNN-SVM-FC. Multiscale detection windows are utilized to adaptively identify various pressure features during the stall process. Additionally, to reduce the false alarm rate, a fuzzy control algorithm is integrated with the temporal accumulation of prediction results from the multi-branch network for joint analysis. A series of test data from a five-stage axial compressor at different operating speeds is used to verify this method. The results indicate that the proposed Multiscale CNN-SVM-FC method enhances the accuracy of classification and reduces the false alarm rate compared to the standard CNN-SVM model, achieving over 99% accuracy in identifying unstable states under various speeds. Compared to three traditional stall prewarning methods, the Multiscale CNN-SVM-FC model provides an average warning signal 164 milliseconds ahead of stall, and reduces the uncertainty associated with threshold selection, which typically relies on engineering experience.

The Evolution of the International Monetary System: Impacts on Multinational Firms and Global Trade

The International Monetary System is a global system that regulates money transfers between different nations. It includes currency exchange rates, regulation of international payments, legislations on capital movements, and worldwide financial organizations, including the IMF. The paper analyses how the change over time of the International Monetary System (IMS) affects the international trade and MNCs. Thus, from the Bretton Woods new exchange regime established in 1944–1971 to the current floating exchange rate system, it has been analyzed in regard to the IMS's historical development. Because of the increasing volatility in the currency rate, this has consequently had an impact on the stability of international transactions and trade volumes. The IMS study investigates exchange rate systems, capital flow regulations, and the role of international financial organizations, including the IMF. For its methodology, the study resorted to a mixed-method approach, combining numbers analysis-that includes regression models and time- series data on exchange rates, amounts of international trade, and MNC performance-with analysis of the historical record and inspection of IMF policy documents. The study shows that trade goes down during big financial troubles, like the Asian Financial Crisis in 1997 and the Global Financial Crisis in 2008. It also shows that changes in exchange rates make global trade smaller, especially during these crises. Also, while IMF help, such as structural adjustment programs, often causes short-term problems for multinational companies (MNCs), like less foreign direct investment (FDI), these policies usually help stabilize economies and lead to recovery in the long run. The outcomes indicate how the IMS, international trade, and multinational corporations are intertwined. Acquainting oneself with the fluctuating money system is very crucial because it enables a person to make informed decisions about the world economy. Keywords: International Monetary System (IMS), Exchange Rate Volatility, Multinational Corporations (MNCs), International Trade, International Monetary Fund (IMF)

Environmental Limitations of the Use of Solar Energy in the Carpathian Region

Abstract. The purpose of the work is to study the environmental limitations of using solar energy in the Carpathian region of Ukraine. Problem. A number of main restrictions affecting the location of solar power plants are highlighted, which are analyzed in the process of carrying out a strategic environmental assessment and are substantiated by calculation in the process of the environmental impact assessment procedure. For solar power plant projects, the main environmental limitation is the use of land areas. In the Carpathian region, plots under solar power plants must bypass agricultural land, recreational and touristic use, buffer zones of nature reserves. An important technical limitation is the ability to connect solar power plants to electrical networks. Methodology. In the course of the research, the regression analysis of the dependence of the amount of electricity produced by the experimental solar power plant at the Ivano-Frankivsk National Technical University of Oil and Gas on meteorological parameters was further developed using statistical processing of a series of observations. Results. Originality. It was determined that the air temperature has the greatest influence on the production of electrical energy with a correlation coefficient of 0.919. A direct linear relationship indicates an increase in power generation with an increase in air temperature up to +250C. Relative humidity has a fairly significant effect on the generation of solar energy. The resulting correlation coefficient of 0.776 indicates a close relationship between the two data series. When the humidity increases to 76%, the production of electricity by solar photovoltaic panels increases and decreases with a further increase in humidity from 75 to 100 %, which is associated with an increase in cloudiness under such weather conditions. Wind speed also has a strong correlation with solar power generation. Electricity production increases in conditions of low wind or almost windless weather with a wind speed value of up to 2.9 m/s and decreases with a further increase in wind speed. Practical value. The obtained significant regularities will be of practical importance for the justification of environmental restrictions when using solar energy for the conditions of the city of Ivano-Frankivsk.

EELS hyperspectral images unmixing - using autoencoders

Spatially resolved Electron Energy-Loss Spectroscopy conducted in a Scanning Transmission Electron Microscope enables the acquisition of hyperspectral images. Spectral unmixing is the process of decomposing each spectrum of an hyperspectral image into a combination of representative spectra (endmembers) corresponding to compounds present in the sample along with their local proportions (abundances). Spectral unmixing is a complex task, and various methods have been developed in different communities using hyperspectral images. However, none of these methods fully satisfy the spatially resolved Electron Energy-Loss Spectroscopy requirements. Recent advancements in remote sensing, which focus on Deep Learning techniques, have the potential to meet these requirements, particularly Autoencoders. In this study, the performance of Deep Learning methods using autoencoders for spectral unmixing is evaluated, and their results are compared with traditional methods. Synthetic hyperspectral images have been created to quantitatively assess the outcomes of the unmixing process using specific metrics. The methods are subsequently applied to a series of experimental data. The findings demonstrate the promising potential of autoencoders as a tool for Electron Energy-Loss Spectroscopy hyperspectral images unmixing, marking a starting point for exploring more sophisticated Neural Networks.

No need for meat as most customers do not leave canteens on Veggie Days

Switching to a diet lower in red meat has the potential to reduce greenhouse gas emissions. Using a unique time series of daily sales data from three German university canteens from 2017 to 2019, we analyse the effects of a monthly Veggie Day in a food-away-from-home context. We find that the temporary ban on meat dishes did not lead to a widespread boycott – as the heated public debates might have suggested. In our setting, a Veggie Day could reduce greenhouse gas emissions by up to 66%. However, especially at the site with a higher share of meat eaters on regular days, up to 22% of customers bypassed the meat-free main dishes on Veggie Days and ate at other on-site alternatives where meat was available. However, total on-site sales did not decrease significantly. Students were less likely to switch to alternatives than staff and guests. A less stringent implementation of a Veggie Day where only beef dishes were removed from the menu, did not result in a significant shift to alternatives but could reduce emissions by up to 51%.

Shanghai Transport Carbon Emission Forecasting Study Based on CEEMD-IWOA-KELM Model

In the light of the worsening of, and the adverse effects produced by, global warming, a study of Shanghai’s transport carbon emissions can provide an advanced model that can be replicated throughout other cities, thus assisting in the management and reduction of carbon emissions. Considering the volatility and nonlinearity of the carbon emission data series of the transport industry, a prediction model combining complementary ensemble empirical modal decomposition (CEEMD), the improved whale optimization algorithm (IWOA), and the Kernel Extreme Learning Machine (KELM) is proposed for a more accurate prediction of the forecasting of carbon emissions from Shanghai’s transport sector. First, nine indicators were screened as the influencing factors of Shanghai’s transport carbon emissions through the STIRPAT model, and the corresponding carbon emissions were calculated with data related to Shanghai’s transport carbon emissions from 1995 to 2019; Secondly, CEEMD was used to decompose the original data into multiple smooth series and one residual term, and KELM was applied to build a prediction model for each decomposition result, and IWOA was used to optimize the model parameters. The experimental results also demonstrate that CEEMD can effectively reduce model errors. Comparative experiments show that the IWOA algorithm can significantly enhance the stability of machine learning models. The outcomes of various experiments indicate that the CEEMD-IWOA-KELM model produces optimal results with the highest accuracy. Additionally, this model exhibits high stability, as it provides a wider range of methods for predicting carbon emissions and contributing to carbon reduction targets.

Detection of aberration in human behavior using shallow neural network over smartphone inertial sensors data

AbstractThe integration of different Mobile Edge Computing (MEC) applications has significantly enhanced the realm of security and surveillance, with Human Activity Recognition (HAR) standing out as a crucial application. The diverse sensors found in smartphones have made it convenient for monitoring applications to gather and analyze data, rendering them valuable for HAR purposes. Moreover, MEC can be employed to automate surveillance, allowing intelligent monitoring of restricted areas to identify and respond to unwanted or suspicious activities. This research develops a system using motion sensors in smartphones to identify unusual human activities. People's smartphones were employed to monitor both suspicious and regular activities. Information was collected for various actions categorized as either suspicious or regular. When a person performs a certain action, the smartphone records a series of sensory data, analyse important patterns from the basic data, and then determines what the person is doing by combining information from different sensors. To prepare the data, information from different sensors was aligned to a shared timeline. In this study, we used a sliding window approach on synchronized data to feed sequences into LSTM and CNN models. These models, which include initial layers of LSTM and CNN, automatically find important patterns in the order of human activities. We combined SVM with the features extracted by the shallow Neural Network to make a mixed model that predicts suspicious activities. Lastly, we compared LSTM, CNN, and our new shallow mixed neural network using a new real‐time dataset. The mixed model of CNN and SVM achieved an accuracy of 94.43%. Additionally, the sliding window method's effectiveness was confirmed with a 4.28% improvement in accuracy.

Weaknesses of steady state analysis when used for selecting solar air heaters

The paper highlights the constrains faced by the user when choosing the most appropriate solar collector from a set of collectors belonging to the same category. Two basic kinds of one porous and one non-porous with single flow/pass flat solar air heaters were tested. The collectors have the same size but different internal shapes and absorber materials. The two collectors were tested and a shading procedure was implemented in order to emphasize several aspects related to the thermal inertia of the collectors. Comparison has been made between the efficiency values obtained experimentally. The shade test shows a faster decrease of the outlet air temperature for V-porous absorber collector in comparison with the U-corrugated absorber collector i.e. 6 °C and –3.5 °C respectively in a time interval of 74 s. Exposing the collectors on a fresh heating cycle under a low stability radiative regime (450–750 W/m2) one observe that 60% of the first 180 s the corrugated collector outperforms the porous collector; after that interval of time the porous collector becomes more effective. The low thermal inertia of the solar air collectors yields high variation of the thermal efficiency in days with low stability of the radiative regime. In such days, using measured series of solar irradiance data with long sampling period between two successive measurements does not allow a proper choice of the most appropriate solar collector type.