Lower Root Mean Square Error Values Research Articles

INFLUENCE OF POINT SOURCES OF POLLUTION ON AIR QUALITY IN MAOPOLSKA FIRST TESTS OF A NEW VERSION OF FORECASTING OF AIR POLLUTION PROPAGATION SYSTEM

Polluted air is dangerous to human life and health. Particulate matter, among others PM10, isone of the most harmful substances. In Maopolska and its capital Krakw, the concentrations ofharmful substances often exceed the standards set by the World Health Organization. Krakw,thanks to the ban on residential heating with solid fuels, has significantly reduced emissions ofpollution, but they remain high in the remaining part of the region, affecting air quality in thecapital as well. With the frequent occurrence of high concentrations of pollutants, in addition tothe necessary measures aimed at reducing emissions, forecasting of air pollutant concentrations isneeded to inform the population if normative concentrations are likely to be exceeded. The FAPPS(Forecasting of Air Pollution Propagation System), based on the AROME/MM5/CALMET/CALPUFF model ensemble, has been operating in Maopolska since 2014 and has been used tocreate pollution concentration forecasts for Krakw and Maopolska. In this study, the influence ofemissions from point sources on air quality in Maopolska was investigated based on the results ofmodellingwith this system. Modelling results indicate that this influence is negligible. The qualityof PM10 forecasts for four versions of the FAPPS system, differing in the meteorological modelused MM5 (Fifth-Generation Penn State/NCAR Mesoscale Model) or WRF (Weather Researchand Forecasting) and the emission input (2015 emissions from the Maopolska Marshals Office,updated with data from the City of Krakw for 2018, or the 2020 emission inventory from theCentral Emission Database) was tested. The quality of forecasts was assessed based on the resultsof measurements at the 22 air pollution monitoring stations for three smog episodes that occurredon 1117.11.2021, 1115.12.2021 and 1318.03.2022. The best results for Krakw were obtainedusing an approach based on the WRF model and the emission inventory from the Central EmissionDatabase, for which an RMSE (Root Mean Square Error) value of 30.02 g/m3 was obtained forselected episodes. In the case of Maopolska, the lowest RMSE value (33.58 g/m3) was obtainedfor the system using the emission inventory obtained from the Marshals Office and the WRF model. First tests indicate that changing the meteorological model from MM5 to WRF can lead toimproved modelling results, but, further research is needed to confirm it.

Neural network model for predicting the biomethane yield in an anaerobic digester using biomass composition profiles

The production of biomethane or renewable natural gas from anaerobic digestion (AD) of agriculture and agroindustry wastes has gained increasing interest towards decarbonisation of fuels and energy supply. Methane yield is a critical step for assessing biomethane projects. In this work, an artificial neural network (ANN) model was developed to predict the methane yield using a collected dataset of 340 experimental data. The input parameters to the network were biomass composition in terms of total solids (TS), volatile solids (VS), lipids, protein, and lignin. The root mean square error (RMSE) of the model was 0.031 L CH4/g VS, with a mean absolute percentage error of 9.2%. The performance of the ANN models was validated using additional mono- and co-digestion experimental datasets. ANN-prediction power was also compared to the outputs of five multiple linear regression models from the literature. Results clearly demonstrated that, compared to the conventional multiple regression models, the proposed ANN-based model has a superior predictive power with lower RMSE values and lower prediction error in most cases. As such, the model is useful for preliminary stages of process design and evaluation of AD-based bioenergy projects. The model has the advantage of requiring only the analysis for five composition values and it has been implemented in a Python function which can also facilitate its wide application.

Experimental investigation and AI prediction modelling of ceramic waste powder concrete – An approach towards sustainable construction

The ceramic waste powder (CWP) is generated in the ceramic industry during the cutting and polishing stages. It is harmful to the environment and needs a massive area for disposal. Therefore, an alternative way is required to reduce the environmental pollution and landfill caused by CWP. The aim of the study is to establish an Artificial Intelligence (AI) model for CWP concrete from the experimental results to save time and cost. Advancements in AI have made the estimation of concrete mechanical characteristics possible by employing Machine Learning (ML) approaches. In the current study, 60 concrete mixes with waste CWP are made as a partial replacement of cement by 10% and 20%. The plain concrete's ultrasonic pulse velocity (UPV) is taken as a reference. Furthermore, supervised ML techniques (i.e., Bagging, XG Boost, AdaBoost) and standalone (Decision tree) are employed to foresee the UPV of CWP concrete (CWPC). The prediction model's performance is evaluated using R2, Root Mean Square Error (RMSE) values, and Mean Absolute Error (MAE). The k-fold cross-validation is used to validate the performance of the prediction model. The XG Boost model, with an R2 value of 0.95, performed better compared to Bagging, AdaBoost, and DT models. Among all ensemble and individual models, the XG Boost model performs better with higher R2 and lower RMSE (0.081 km/s) and MAE (0.063 km/s) values. Therefore, the CWPC, as a construction material, would reduce land degradation and water pollution. In addition, applying ML techniques for estimating concrete characteristics would have reduced the consumption of efforts, resources, and time of researchers in the construction sector.

Prediction of Air Quality Index Using Machine Learning Techniques: A Comparative Analysis

An index for reporting air quality is called the air quality index (AQI). It measures the impact of air pollution on a person’s health over a short period of time. The purpose of the AQI is to educate the public on the negative health effects of local air pollution. The amount of air pollution in Indian cities has significantly increased. There are several ways to create a mathematical formula to determine the air quality index. Numerous studies have found a link between air pollution exposure and adverse health impacts in the population. Data mining techniques are one of the most interesting approaches to forecast AQI and analyze it. The aim of this paper is to find the most effective way for AQI prediction to assist in climate control. The most effective method can be improved upon to find the most optimal solution. Hence, the work in this paper involves intensive research and the addition of novel techniques such as SMOTE to make sure that the best possible solution to the air quality problem is obtained. Another important goal is to demonstrate and display the exact metrics involved in our work in such a way that it is educational and insightful and hence provides proper comparisons and assists future researchers. In the proposed work, three distinct methods—support vector regression (SVR), random forest regression (RFR), and CatBoost regression (CR)—have been utilized to determine the AQI of New Delhi, Bangalore, Kolkata, and Hyderabad. After comparing the results of imbalanced datasets, it was found that random forest regression provides the lowest root mean square error (RMSE) values in Bangalore (0.5674), Kolkata (0.1403), and Hyderabad (0.3826), as well as higher accuracy compared to SVR and CatBoost regression for Kolkata (90.9700%) and Hyderabad (78.3672%), while CatBoost regression provides the lowest RMSE value in New Delhi (0.2792) and the highest accuracy is obtained for New Delhi (79.8622%) and Bangalore (68.6860%). Regarding the dataset that was subjected to the synthetic minority oversampling technique (SMOTE) algorithm, it is noted that random forest regression provides the lowest RMSE values in Kolkata (0.0988) and Hyderabad (0.0628) and higher accuracies are obtained for Kolkata (93.7438%) and Hyderabad (97.6080%) in comparison to SVR and CatBoost regression, whereas CatBoost regression provides the highest accuracies for New Delhi (85.0847%) and Bangalore (90.3071%). This demonstrated definitely that datasets that had the SMOTE algorithm applied to them produced a higher accuracy. The novelty of this paper lies in the fact that the best regression models have been picked through thorough research by analyzing their accuracies. Moreover, unlike most related papers, dataset balancing is carried out through SMOTE. Moreover, all of the implementations have been documented via graphs and metrics, which clearly show the contrast in results and help show what actually caused the improvement in accuracy.

Comparative performance of Sentinel-2 MSI and Landsat-8 OLI data in canopy cover prediction using Random Forest model: Comparing model performance and tuning parameters

Quantifying canopy cover using Random Forest (RF) model’s appropriate tuning parameters value and sensor based predictor variables is always challenging, especially in fragmented dry deciduous forests. Therefore, this study was designed to compare the performances of Sentinel-2 and Landsat-8 based models using the RF model for predicting canopy cover with assessing variables’ relative importance and correlation. Sentinel-2 and Landsat-8 based bands and spectral indices were used as predictor variables. We compared different mtry, ntree and bag fraction values of the RF model. R-square (R2) and root mean square error (RMSE) were used for comparing the model performance. The results showed that the lowest RMSE value was associated with the default value (predictors/3) or more than the default value of mtry, with bag fraction 0.3–0.7 for Sentinel-2 and 0.3–0.4 for Landsat-8. Model accuracy has increased and stabilized with increase of ntree, and received the lowest RMSE to ntree of more than 1000. Except for SWIR indices based model of Landsat-8, all other Landsat-8 based model’s accuracy was lesser compared to Sentinel-2 based models. Model accuracy of Sentinel-2 based full model (except red edge indices) was marginally better (R2 = 0.899, RMSE = 6.883 %) than Landsat-8 based full model (R2 = 0.886, RMSE = 7.089 %). But with the incorporation of red edge indices, full model RMSE had decreased further from 6.883 % to 6.747 %, and R2 had increased from 0.899 to 0.918. The full model of Sentinel-2 tended to spread variable importance among more variables, but the full model of Landsat-8 slightly tends to concentrate variable importance with fewer variables. However, SWIR bands and indices were the most important predictor variables and highly correlated with canopy cover. These findings can solve the parameter value choice of RF model, and the use of the Sentinel-2 based model will be superior to Landsat-8 based model.

Deep learning-based marine big data fusion for ocean environment monitoring: Towards shape optimization and salient objects detection

ObjectiveDuring the last few years, underwater object detection and marine resource utilization have gained significant attention from researchers and become active research hotspots in underwater image processing and analysis domains. This research study presents a data fusion-based method for underwater salient object detection and ocean environment monitoring by utilizing a deep model.MethodologyA hybrid model consists of an upgraded AlexNet with Inception v-4 for salient object detection and ocean environment monitoring. For the categorization of spatial data, AlexNet is utilized, whereas Inception V-4 is employed for temporal data (environment monitoring). Moreover, we used preprocessing techniques before the classification task for underwater image enhancement, segmentation, noise and fog removal, restoration, and color constancy.ConclusionThe Real-Time Underwater Image Enhancement (RUIE) dataset and the Marine Underwater Environment Database (MUED) dataset are used in this research project’s data fusion and experimental activities, respectively. Root mean square error (RMSE), computing usage, and accuracy are used to construct the model’s simulation results. The suggested model’s relevance form optimization and conspicuous item prediction issues in the seas is illustrated by the greatest accuracy of 95.7% and low RMSE value of 49 when compared to other baseline models.

Forecasting of groundnut yield using meteorological variables

Agriculture is one of the most significant economic sectors in India, which is strongly dependent on climatic conditions. Groundnut is one of the major oilseed crops of Gujarat and Junagadh has highest production of groundnut in Gujarat. For effective, forward-looking, and current planning, especially in agriculture, which is fraught with uncertainty, reliable and timely forecasts are essential. Therefore, effective yield forecast of such important oilseed crop is necessary for future planning and policy making. In the present investigation, the time series data of groundnut yield and weather parameters of 29 years Junagadh (1991-92 to 2019-20) were used. The week wise weather indices were generated using correlation between de-trend yield and weekly weather variables. Multiple Linear Regression (MLR) and Discriminant Function Analysis were used to develop yield forecasting model using weather indices before three week and one week before harvest, respectively. These models were compared using Coefficient of multiple determination (Adj. R2), Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The study found that model-1A, developed using MLR technique have high value of adj. R2(78.8%) and low RMSE value (671.72). Multiple linear regression (MLR) was found to be more accurate than the discriminant function analysis approach.

Mini plant factory development using IoT and cloud system for urban greens cultivation

A significant urban population increases with the rising demand for higher food quality and security. Adapting the current food chain system and our eating patterns are needed to make them more sustainable. Plant factories are one approach for making the food chain system more sustainable. However, plant factory systems typically require significant initial capital expenditure. As a result, a mini plant factory, which is more affordable and flexible, will be appropriate in urban areas. The objective of this study was to design, build, and evaluate a mini plant factory system for horticultural crop production. The mini plant factory system comprises a mini cultivation room with three growing shelves equipped with a microclimate sensor, nutrient dosing system, artificial growth light, and ventilation system. Sensors and actuators are integrated with the Agrieye Cloud system for monitoring and control and can autonomously execute data retrieval and actuation. Mini Plant Factory operates by automatically capturing monitoring data every five minutes; if a value exceeds the limit, the actuator control system activates to stabilize that value. The linear regression test yielded the most excellent R2 value of 0.999 for sensor calibration. The validation test was conducted using RMSE (Root Mean Square Error) and MAPE (Mean Absolute Percentage Error). The results show that the lowest RMSE values are 0.133, and the lowest MAPE values are 0.082%. The data logging system’s performance is perfect. The control and scheduling system works well but does not function optimally since there are no interruptions throughout the data gathering process.

The use of artificial neural network (ANN) in dry flue gas desulphurization modelling: Levenberg–Marquardt (LM) and Bayesian regularization (BR) algorithm comparison

AbstractThis research project aims to investigate the efficacy of artificial neural networks (ANN) in mapping dry flue gas desulphurization (DFGD). Bayesian regularization (BR) and Levenberg–Marquardt (LM) training algorithms were used for DFGD modelling. The input layer feed data contained diatomite to Ca(OH)2 ratio, hydration time, hydration temperature, sulphation temperature, and inlet gas concentration, while the output layer metadata were sorbent conversion and sulphation responses. The hyperbolic tangent (tansig), sigmoid (logsig), and linear (purelin) activation functions were compared to ascertain the best network learning model. The number of hidden layer cells also varied between 7 and 10, given the existence of multiple output feed data. BR and LM performance evaluation was based on coefficient of determination (R2), root mean square error (RMSE), and mean square error (MSE) mathematical analysis. BR was a superlative training tool compared to LM, with lower RMSE and MSE values. The goodness of fit data for both techniques was close to unity, clarifying that ANN using BR and LM tools can be used to predict DGFD outcome. The shrinking core model was used to analyze the desulphurization reaction and concluding the chemical reaction was the reaction controlling step.

Parameter extraction for photovoltaic models with tree seed algorithm

Evaluation, simulation and optimization of PV systems is very important for fast and accurate parameter extraction based on current–voltage and power–voltage characteristic curves of photovoltaic models. Therefore, researchers used many metaheuristic algorithms to estimate the parameters of various PV modules. In this study, tree seed algorithm (TSA) was used for parameter estimation of the STM6-40/36 PV module. In the basic TSA, there are two different resolution mechanisms for balancing both local and global search technique. For this reason, basic TSA was preferred for parameter extraction of the PV module. The parameter results obtained by TSA were compared with those found by some other algorithms in the literature. According to the comparison result, the lowest root mean square error (RMSE) was obtained with the TSA algorithm. When the convergence graphs are examined, it is seen that TSA converges faster than other algorithms. When the box plots are analyzed, the results obtained by TSA have fewer outliers than the results of other algorithms, showing that TSA has a stable structure. In addition, a ranking graph was drawn according to the results obtained by all algorithms at each run time, and it was seen that TSA had the lowest RMSE value at all run times. Thus, it is concluded that TSA is a very competitive method for the PV module problem.

Application of Six Metaheuristic Optimization Algorithms and Random Forest in the uniaxial compressive strength of rock prediction

The uniaxial compressive strength (UCS) is one of the most important parameters for judging the mechanical behaviour of rock mass in rock engineering design and excavation such as tunnels, subways, drilling, slopes and mines stability. However, an obvious deficiency of traditional experimental operations to obtain UCS is that it suffers from a lack of efficiency and accuracy. Therefore, the prediction of the UCS of rock is of high practical significance in reducing evaluation time and improving the precision of results. At the same time, breaking the universality problem of traditional empirical models and improving the accuracy of artificial intelligence models need to absorb and accommodate more rock samples. Hence, a total of 226 rock samples with five properties were carried out from four published studies and selected to generate a dataset in this investigation, i.e., Granitic, Caliche, Schist, Sandstone and Grade III granitic. Five individual parameters of rock samples consisting of Schmidt hardness rebound number (SHR), P-wave velocity (Vp), point load strength (Is(50)), porosity (Pn), and density (D) were used to predict UCS. In this paper, six metaheuristic optimization algorithms were utilized to improve the performance of the Random Forest (RF) model, i.e., slime mould algorithm (SMA), chameleon swarm algorithm (CSA), transient search optimization (TSO), equilibrium optimizer (EO), social network search (SNS) and student psychology based optimization algorithm (SPBO). Four performance indices , the root mean square error (RMSE), the determination coefficient (R2), Willmott’s index (WI) and the variance accounted for (VAF) were utilized to evaluate the performance of all models in forecasting the UCS of rock. The results of the performance comparison demonstrated that the TSO-RF model has the highest values of R2 (train: 0.9923 and test: 0.9753), WI (train: 0.9980 and test: 0.9937), and VAF (train: 99.2272% and test: 97.6852%), the lowest values of RMSE (train: 3.8313 and test: 6.5968) compared to the other models. The research in this study provided an effective attempt to further improve the accuracy of UCS prediction.

A Comparison of the Pearson, Spearman Rank and Kendall Tau Correlation Coefficients Using Quantitative Variables

In all fields and branches of sciences especially statistics, the correlation coefficient is one of the most often used statistical measures. This study has been carried out for comparing the performances of the Pearson (), Spearman's Rank (), and Kendall’s Tau () correlation coefficients under three sample sizes based on the data of quantitative variables of cotton. Descriptive statistics showed the presence of genetic variability for the cotton studied traits in this study. The quantity, significance, and direction of the correlation calculated by differed in some cases from the other methods under the three sample sizes, opposite is true for and . The highest number of positive correlations among studied traits were by under N = 30 observations, and by and under N = 20 observations. The studied correlation methods performances by Root Mean Square Error (RMSE) revealed that and appear to be a good estimator of correlation because they have the lowest values of RMSE. The highest values of RMSE were observed by and under N=10 and N=20, and by under N=30. The results of PCA could be useful and appropriate in this study, in which the PCA1 had highly positively correlated with the three studied methods for N=10 observations, and with and for N=20 observations.

Thermal analysis of concrete slabs with insulating materials using ANSYS

PurposeThis study aims to assess the efficacy of thermal analysis of concrete slabs by including different insulation materials using ANSYS. Regression equations were proposed to predict the thermal conductivity using concrete density. As these simulation and regression analyses are essential tools in designing the thermal insulation concretes with various densities, they sequentially reduce the associated time, effort and cost.Design/methodology/approachTwo grades of concretes were taken for thermal analysis. They were designed by replacing the natural fine aggregates with thermal insulation aggregates: expanded polystyrene, exfoliated vermiculite and light expanded clay. Density, temperature difference, specific heat capacity, thermal conductivity and time were measured by conducting experiments. This data was used to simulate concrete slabs in ANSYS. Regression analysis was performed to obtain the relation between density and thermal conductivity. Finally, the quality of the predicted regression equations was assessed using root mean square error (RMSE), mean absolute error (MAE), integral absolute error (IAE) and normal efficiency (NE).FindingsANSYS analysis on concrete slabs accurately estimates the thermal behavior of concrete, with lesser error value ranges between 0.19 and 7.92%. Further, the developed regression equations proved accurate with lower values of RMSE (0.013 to 0.089), MAE (0.009 to 0.088); IAE (0.216 to 5.828%) and higher values of NE (94.16 to 99.97%).Originality/valueThe thermal analysis accurately simulates the experimental transfer of heat across the concrete slab. Obtained regression equations proved helpful while designing the thermal insulation concrete.

The Applicability of Biogeography-Based Optimization and Earthworm Optimization Algorithm Hybridized with ANFIS as Reliable Solutions in Estimation of Cooling Load in Buildings

The foundation of energy-efficient architectural design is modeling heating and cooling loads (HLs and CLs), which defines the heating and cooling apparatus constraints necessary to maintain a suitable interior air environment. It is possible that analytical models for energy-efficient buildings might offer an accurate evaluation of the influence that various building designs would have. The implementation of these instruments, however, might be a process that requires a significant amount of manual labor, a significant amount of time, and is reliant on user experiences. In light of this, the authors of this paper present two unique methods for estimating the CL of residential structures in the form of complex mathematical concepts. These methodologies include an evolutionary web algorithm (EWA), biogeography-based optimization (BBO), and a hybridization of an adaptive neuro-fuzzy interface system (ANFIS), namely BBO-ANFIS and EWA-ANFIS. The findings initiated from each of the suggested models are evaluated with the help of various performance metrics. Moreover, it is possible to determine which model is the most effective by comparing their coefficient of determination (R2 ) and its root mean square error (RMSE) to each other. In mapping non-linear connections between input and output variables, the observed findings showed that the models used have a great capability. In addition, the results showed that BBO-ANFIS was the superior forecasting model out of the two provided models, with the lowest value of RMSE and the greatest value of R2 (RMSE = 0.10731 and 0.11282 and R2 = 0.97776 and 0.97552 for training and testing phases, respectively). The EWA-ANFIS also demonstrated RMSE and R2 values of 0.18682 and 0.17681 and 0.93096 and 0.93874 for the training and testing phases, respectively. Finally, this study has proven that ANN is a powerful tool and will be useful for predicting the CL in residential buildings.

A regional machine learning method to outperform temperature-based reference evapotranspiration estimations in Southern Spain

Accurate estimations of reference evapotranspiration (ET0) are crucial for determining crop water requirements and designing an adequate irrigation scheduling to optimize the use of water. In this work, a new clustering method to outperform the accuracy of ET0 estimations only using temperature variables has been developed and assessed, based on the multifractal properties of the Diurnal Temperature Range (DTR). Thus, a more accurate weather stations’ grouping method has been evaluated, regardless of their geographic location. All the datasets were collected from 89 automated weather stations in the period 2000–2018 and pooled into two main regions (1 and 2). In each region, an iterative procedure has been carried out: 1) selection of all the stations except the candidate one for the training procedure and 2) test procedure using the candidate station. The results showed that Machine Learning models (ML) highly outperformed Hargreaves-Samani (HS) in most of the cases, being Multilayer Perceptron (MLP) the most accurate over Extreme Learning Machine models (ELM). On average, the results obtained by MLP using the best configuration in the first region were better than those obtained in the second region. Specifically, the first region got an Root Mean Square Error (RMSE) = 0.657 mm/d, Nash–Sutcliffe Efficiency (NSE) = 0.897, Coefficient of Determination (R2) = 0.931 and Mean Bias Error (MBE) = |0.04|mm/d while the second region obtained an RMSE = 0.703 mm/d, NSE = 0.867, R2 = 0.897 and MBE = |0.045|mm/d. Regarding the seasonal performance, spring and autumn obtained the best NSE and R2 results, whereas winter carried out the lowest RMSE values. This study provides a new and more accurate methodology to improve ET0 estimations on a regional basis and only using temperature data in the whole process.

Land Price Classification Map in Jakarta Using Random Forest and Ordinary Kriging

This research provides information about land prices in Jakarta by classifying using the Random Forest method. Where Random Forest is a data mining technique that is usually used to perform classification and regression. Random Forest is one of the best classification methods. It is found that classification accuracy will increase dramatically as a result of voting to select class types and ensemble tree growth. The method helps in providing information about the classification of land prices with the class of land prices per meter less than IDR 15 million, land prices per meter with a price range of IDR 15 to 25 million and land prices per meter more than IDR 25 million. With a fairly good accuracy of 82%, this method can classify where the permeter land price data that is tested will match the predicted classification accurately. Classification is performed on unbalanced data which is then oversampled using the ADASYN method. Assisted by doing spatial interpolation with the Ordinary Kriging method using Semivariogram, information about the classification of land prices can be seen on the distribution of the Jakarta area map. Ordinary Kriging can predict the estimated price per meter of land around the area of land that has a known price. The Root Mean Square Error (RMSE) results of the best Semivariogram model are obtained from the lowest RMSE value, namely the Spherical model with a value of 1.014896e7. The contribution of this research is to provide information about a reliable classification method, namely Random Forest and Ordinary Kriging performance as a spatial analysis method that can predict land prices per meter at unknown points so as to provide information about the distribution of land prices in Jakarta with each price class.

Horizontal and Vertical Geometric Accuracy of Agisoft Photoscan and Pix4D Mapper Softwares at Kebun Raya Universitas Mulawarman in Samarinda, East Kalimantan, Indonesia

The advent of unmanned aerial vehicles has improved aerial photography, which is now aided by computer software for data acquisition and processing. These developments allow for accurate aerial photographs of Kebun Raya Universitas Mulamarwan Samarinda (KRUS) for managerial purposes. This study aims to compare the accuracy of horizontal and vertical geometry with Agisoft Photoscan (AP) and Pix4D Mapper (PM) software. The materials consisted of 150 aerial photographs of KRUS with eight ground control points (GCP) and one internal control point (ICP). Data were obtained from nine flight paths with AP and PM software linked to GCP and ICP. These data were processed and compared to manual measurements using linear error (LE) 90 and circular error (CE) 90 criteria. The commission omission equation was used for the object accuracy test on orthophoto. The vertical geometry accuracy test was carried out using the criteria of root mean square error (RMSE) and LE90 values on the digital elevation model. In comparison, the horizontal geometry accuracy test was performed by using the obtained orthophoto based on RMSE and CE90 criteria. The result demonstrated that the horizontal geometry accuracy for AP was higher than PM, indicated by lower RMSE and CE90 values, which were 0.091 versus 0.148 and 0.139 versus 0.224 for AP and PM, respectively. The vertical geometry accuracy was similar, with lower RMSE and LE90 values of 0.169 versus 0.309 and LE90 0.279 versus 0.224 for AP and PM, respectively. Furthermore, AP is a potential tool for KRUS spatial mapping.

Assessment of Global Geopotential Models for Modelling Malaysia Marine Geoid

The evaluation towards global geopotential models represents a significant part in modelling the localised Marine Geoid. The marine geoid provides the vertical reference information in Marine Spatial Data Infrastructures (MSDI) development response to United Nations Sustainable Development Goals 14 for the sustainable development in marine environment. The main purpose of this study is to select the best model from both combined missions and satellite-only missions for the Malaysian region. The gravity anomaly field from 30 global models were exclusively calculated over the selected study area within 11 years period-time. Afterwards, each dataset was extracted from the ICGEM server to evaluate with the airborne-derived gravity anomaly from the Department of Surveying and Mapping, Malaysia. The internal accuracy, root mean square error (RMSE) and differences between every model and airborne data were computed. The result indicates GGM-derived gravity anomaly for the best combined mission is GECO with RMSE of 8.44 mGal and the standard deviation value of 28.034 mGal. While, the model from Gravity field and steady state Ocean Circulation Explorer (GOCE) namely, the GO_CONS_GCF_2_DIR_R5 is the best for the satellite-only mission with RMSE of 17.43 mGal and the standard deviation value of 22.828 mGal. As a conclusion, GECO model is preferred as the best fit for determining the marine geoid as it has the lowest RMSE value between both mission and the maximum degree of 2109ocoverage. The finding can assist in development of marine geoid for modelling precise surface elevation.

Statistical Evaluation and Trend Analysis of ANN Based Satellite Products (PERSIANN) for the Kelani River Basin, Sri Lanka

Satellite-based precipitation products, (SbPPs) have piqued the interest of a number of researchers as a reliable replacement for observed rainfall data which often have limited time spans and missing days. The SbPPs possess certain uncertainties, thus, they cannot be directly used without comparing against observed rainfall data prior to use. The Kelani river basin is Sri Lanka’s fourth longest river and the main source of water for almost 5 million people. Therefore, this research study aims to identify the potential of using SbPPs as a different method to measure rain besides using a rain gauge. Furthermore, the aim of the work is to examine the trends in precipitation products in the Kelani river basin. Three SbPPs, precipitation estimation using remotely sensed information using artificial neural networks (PERSIANN), PERSIANN-cloud classification system (CCS), and PERSIANN-climate data record (CDR) and ground observed rain gauge daily rainfall data at nine locations were used for the analysis. Four continuous evaluation indices, namely, root mean square error (RMSE), (percent bias) PBias, correlation coefficient (CC), and Nash‒Sutcliffe efficiency (NSE) were used to determine the accuracy by comparing against observed rainfall data. Four categorical indices including probability of detection (POD), false alarm ratio (FAR), critical success index (CSI), and proportional constant (PC) were used to evaluate the rainfall detection capability of SbPPs. Mann‒Kendall test and Sen’s slope estimator were used to identifying whether a trend was present while the magnitudes of these were calculated by Sen’s slope. PERSIANN-CDR performed well by showing better performance in both POD and CSI. When compared to observed rainfall data, the PERSIANN product had the lowest RMSE value, while all products indicated underestimations. The CC and NSE of all three products with observed rainfall data were also low. Mixed results were obtained for the trend analysis as well. The overall results showed that all three products are not a better choice for the chosen study area.

Correlation Analysis of Factors Affecting Firm Performance and Employees Wellbeing: Application of Advanced Machine Learning Analysis

Given the importance of identifying key performance points in organizations, this research intends to determine the most critical intra- and extra-organizational elements in assessing the performance of firms using the European Company Survey (ECS) 2019 framework. The ECS 2019 survey data were used to train an artificial neural network optimized using an imperialist competitive algorithm (ANN-ICA) to forecast business performance and employee wellbeing. In order to assess the correctness of the model, root mean square error (RMSE), mean absolute percentage error (MAPE), mean square error (MSE), correlation coefficient (r), and determination coefficient (R2) have been employed. The mean values of the performance criteria for the impact of internal and external factors on firm performance were 1.06, 0.002, 0.041, 0.9, and 0.83, and the value of the performance metrics for the impact of internal and external factors on employee wellbeing were 0.84, 0.0019, 0.0319, 0.83, and 0.71 (respectively, for MAPE, MSE, RMSE, r, and R2). The great performance of the ANN-ICA model is indicated by low values of MAPE, MSE, and RMSE, as well as high values of r and R2. The outcomes showed that “skills requirements and skill matching” and “employee voice” are the two factors that matter most in enhancing firm performance and wellbeing.