Different Types Of Regression Models Research Articles

Retrieval performance of mangrove tree heights using multiple machine learning regression models and UAV-LiDAR point clouds

ABSTRACT Mangroves are vital coastal ecosystems that provide crucial links between land and sea. Tree height is a key indicator for assessing mangroves’ health status. Currently, there are still numerous challenges in estimating mangrove tree height. In this study, multiple deep learning and shallow machine learning regression models were developed to accurately estimate mangrove tree height using multi-dimensional Light Detection and Ranging (LiDAR) point clouds and their derivatives. We constructed a novel CNN_RepMLP model for mangrove tree height mapping. We also further verified the applicability of different types of regression models in estimating mangrove tree heights, and explored the influence of different LiDAR-derived features on the inversion accuracy for tree heights. The results indicated the following: (1) The CNN_RepMLP displayed satisfactory performance in the inversion of mangrove tree heights, and exhibited better robustness and generalization ability than the convolutional neural network (CNN) model. (2) Among the different LiDAR-derived feature combinations, combining height variables with intensity variables can not only mitigate the negative impact of height variables on inversion models, but also enhance the mangrove tree height inversion accuracy. (3) The ensemble learning framework with ExtraTrees as the meta-model can make better use of the differences and complementarities between different single base models, and exhibited better accuracy in estimating the height of mangrove trees compared with other ensemble learning models. (4) Multiple machine learning models based on multi-dimensional UAV-LiDAR point-cloud-derived features are suitable for the inversion of mangrove tree height. The CNN_RepMLP model outperformed the CNN and stacking ensemble learning models and had more detailed differentiation in terms of mangrove tree height. Its prediction results can more realistically reflect the spatial differentiation characteristics of mangrove tree height.

Ensemble machine learning algorithm for diabetes prediction in Maiduguri, Borno State

Diabetes mellitus (DM) is a metabolic disease characterised by high levels of glucose in the blood, known as hyperglycemia, that can result in multiple problems within the body. The World Health Organisation (WHO) data for 2021 reveals a substantial increase in the prevalence of diabetes mellitus (DM), with the number of cases rising from 108 million in 1980 to 422 million in 2014. Between 2000 and 2019, there was a 3% increase in mortality rates associated with diabetes, categorised by age. In 2019, DM caused the deaths of more than 2 million people. These concerning figures clearly necessitate an immediate response. An alarming incidence of diabetes among the population of Maiduguri and Borno State inspired this investigation. This research proposed stacking ensemble learning approach to predict the rate of occurrence of diabetes cases in Maiduguri. The paper used different types of regression models to predict the occurrences of diabetes cases in Maiduguri over time. These models included adaptive boosting regression (Adaboost), gradient boosting regression (GBOOST), random forest regression (RFR), ordinary least square regression (OLS), least absolute shrinkage selection operator regression (LASSO), and ridge regression (RIDGE). The performance indicators studied in this work are root mean square (RMSE), mean absolute error (MAE), and mean square error (MSE). These metrics were used to assess the effectiveness of both the machine learning and proposed Stacking Ensemble Learning (SEL) approaches. Performance metrics considered in this study are root mean square (RMSE), mean absolute error (MAE), and mean square error (MSE), which were used to evaluate the performance of the machine learning and the proposed Stacking Ensemble Learning (SEL) technique. Experimental results revealed that SEL is a better predictor compared to other machine learning approaches considered in this work with an RMSE of 0.0493; a MSE of 0.0024; and a MAE of 0.0349. It is hoped that this research will help government officials understand the threat of diabetes and take the necessary mitigation actions.

Enhancing sediment transport predictions through machine learning-based multi-scenario regression models

Machine learning is one effective way of increasing the accuracy of sediment transport models. Machine learning captures patterns in the sequence of both structured and unstructured data and uses it for forecasting. In this research, the different regression models were train to forecast sediment data using 8 years of measured sediment data collected in Sg. Linggui suspended sediment station. Data from different scenarios were used where each scenario indicates the number of lags. Seven regression models, namely, Linear Regression, Regression Trees, Support Vector Machines, Gaussian Process Regression, Kernel Approximation, Ensemble of Trees, and Neural Network were trained using the data and compared. The trained models were evaluated using Root Mean Square Error (RMSE) and Coefficient of Determination (R2). The best-performing models from two different types of regression models were chosen and they were tested using the test data to find the Relative Percentage Error (RPE) of the predicted data. The Exponential Gaussian Process Regression model performs much better than the other models in terms of RMSE and R2 values. When the exponential models from all 3 scenarios are compared, scenario 3 seems to have a better-performing model but only by a very small margin, after using testing data, the result shows scenario 3 has less RPE compared to other models. Hence, it can be deduced that the exponential gaussian process regression model from scenario 3 is the best-performing model overall in terms of RSME, R2, and RPE.

Vehicle Localization in 3D World Coordinates Using Single Camera at Traffic Intersection

Optimizing traffic control systems at traffic intersections can reduce the network-wide fuel consumption, as well as emissions of conventional fuel-powered vehicles. While traffic signals have been controlled based on predetermined schedules, various adaptive signal control systems have recently been developed using advanced sensors such as cameras, radars, and LiDARs. Among these sensors, cameras can provide a cost-effective way to determine the number, location, type, and speed of the vehicles for better-informed decision-making at traffic intersections. In this research, a new approach for accurately determining vehicle locations near traffic intersections using a single camera is presented. For that purpose, a well-known object detection algorithm called YOLO is used to determine vehicle locations in video images captured by a traffic camera. YOLO draws a bounding box around each detected vehicle, and the vehicle location in the image coordinates is converted to the world coordinates using camera calibration data. During this process, a significant error between the center of a vehicle's bounding box and the real center of the vehicle in the world coordinates is generated due to the angled view of the vehicles by a camera installed on a traffic light pole. As a means of mitigating this vehicle localization error, two different types of regression models are trained and applied to the centers of the bounding boxes of the camera-detected vehicles. The accuracy of the proposed approach is validated using both static camera images and live-streamed traffic video. Based on the improved vehicle localization, it is expected that more accurate traffic signal control can be made to improve the overall network-wide energy efficiency and traffic flow at traffic intersections.

EU: The Effect of Energy Factors on Economic Growth

In this article, we investigate the effect of different energy variables on economic growth of several oil-importing EU member states. Three periods from 2000 to 2020 were investigated. Three different types of regression models were constructed via the gretl software. Namely, the OLS, FE, and SE approaches to panel data analysis were investigated. The FE approach was chosen as the final one. The results suggest the importance of the consumption of both oil and renewable energy on economic growth. Crises of certain periods also had a noteworthy effect as well.

Characterization of Traffic Accidents Based on Long-Horizon Aggregated and Disaggregated Data

For sustainable transportation systems, modeling road traffic accidents is essential in order to formulate measures to reduce their harmful impacts on society. This study investigated the outcomes of using different datasets in traffic accident models with a low number of variables that can be easily manipulated by practitioners. Long-horizon aggregated and disaggregated road traffic accident datasets on Egyptian roads (for five years) were used to compare the model’s fit for different data groups. This study analyzed the results of k-means data clustering and classified the data into groups to compare the fit of the base model (Smeed’s model and different types of regression models). The results emphasized that the aggregated data used had less efficiency compared with the disaggregated data. It was found that the classification of the disaggregated dataset into reasonable groups improved the model’s fit. These findings may help in the better utilization of the available road traffic accident data for determining the best-fitting model that can assist decision-makers to choose suitable road traffic accident prevention measures.

Bayesian model predicts the aboveground biomass of Caragana microphylla in sandy lands better than OLS regression models

Abstract Aims In forest ecosystems, different types of regression models have been frequently used for the estimation of aboveground biomass, where Ordinary Least Squares (OLS) regression models are the most common prediction models. Yet, the relative performance of Bayesian and OLS models in predicting aboveground biomass of shrubs, especially multi-stem shrubs, has relatively been less studied in forests. Methods In this study, we developed the biomass prediction models for Caragana microphylla Lam. which is a widely distributed multi-stems shrub, and contributes to the decrease of wind erosion and the fixation of sand dunes in the Horqin Sand Land, one of the largest sand lands in China. We developed six types of formulations under the framework of the regression models, and then, selected the best model based on specific criteria. Consequently, we estimated the parameters of the best model with OLS and Bayesian methods with training and test data under different sample sizes with the bootstrap method. Lastly, we compared the performance of the OLS and Bayesian models in predicting the aboveground biomass of C. microphylla. Important Findings The performance of the allometric equation (power = 1) was best among six types of equations, even though all of those models were significant. The results showed that mean squared error of test data with non-informative prior Bayesian method and the informative prior Bayesian method was lower than with the OLS method. Among the tested predictors (i.e. plant height and basal diameter), we found that basal diameter was not a significant predictor either in OLS or Bayesian methods, indicating that suitable predictors and well-fitted models should be seriously considered. This study highlights that Bayesian methods, the bootstrap method and the type of allometric equation could help to improve the model accuracy in predicting shrub biomass in sandy lands.

Does certification improve hired labour conditions and wageworker conditions at banana plantations?

Certification of banana plantations is widely used as a device for protecting and improving socio-economic conditions of wageworkers, including their incomes, working conditions and—increasingly—voice [related to labour relations and workplace representation]. However, to date, evidence about the effectiveness of certification in these domains is scarce. We collected detailed field data on (1) economic benefits for improving household income, (2) social benefits for labour practices, and (3) the voice of wageworkers focusing on identity and identification issues amongst wageworkers at Fairtrade certified banana plantations and comparable, non-certified plantations in the Dominican Republic. We used different types of regression models to identify significant relationships. Econometrical analysis of survey results complemented by field observations and outcomes from in-depth stakeholder interviews indicate that the impact of Fairtrade certification on wageworkers’ economic benefits is rather limited. However, the impact on the voice of wageworkers (job satisfaction, sense of ownership, trust), is more evident. On Fairtrade certified plantations workers are more satisfied with the course of life and better represented. Thus while the additional value of Fairtrade certification on primary wages seems limited, Fairtrade has relevant positive effects on the labour force, particularly by delivering in-kind benefits, offering a sense of job-security, improving voice and enabling private savings. Benefits of (Fairtrade) certification, but also other interventions with a similar purpose, might therefore not be discerned in terms of economic benefits such as wages or basic labour conditions that are under direct control of (inter)national law, but they should be identified in terms of social benefits and improved norms of conduct for wageworker engagement.

Distribution of Butterfly Species Associated with Environmental Factors in Sri Lanka

The species diversity monitoring of butterflies in Sri Lanka is considered in this study under certain environmental factors. Species richness, and Shannon and Simpson’s diversity indices were calculated to understand the variation of the distributions of butterfly species. Maximum and minimum diversity and richness were observed from Rathnapura and Puththalama districts in Sri Lanka, respectively. Based on the Diamond’s assembly rules and Probabilistic models, it was noted that most of the butterflies were randomly distributed, and there was little predictable co-occurrence between species pairs. To study the distributional patterns of butterfly species with environmental factors, five different types of regression models were fitted by considering the occurrences of each species. The results clearly indicated that the distribution of butterfly species varies from species to species according to the different environmental factors. Further, the occurrence of most of the butterfly species depends on temperature and total rain fall. Prediction of species occurrences with respect to the environmental factors can be done by using the best fitted model of each species. The methodology and results of the study can be adapted to monitor the biodiversity of a certain area.

Sensitivity of extreme rainfall to temperature in semi-arid Mediterranean regions

Warmer air has the potential to hold more water vapour and, therefore, to provide more water to rainfall events. Studying the relationship between rainfall and temperature represents an emerging issue in hydrology and meteorology, since it can be considered fundamental for evaluating the effects of global warming on future precipitation. Various approaches have been tested across different parts of the world, in many cases observing an intensification of extreme precipitation at higher temperatures consistent with the well-known thermodynamic Clausius-Clapeyron relation (CC-scaling rate of 6–7%°C−1). However, at different locations for hourly time-scales, the temperature-extreme rainfall scaling can be higher (super-CC) or lower (sub-CC).This study contributes to the understanding of the scaling relationship between extreme rainfall and temperature under climate conditions characteristic of Mediterranean semi-arid regions, rarely explored in the past. The role of different factors, such as rainfall characteristics and climatic seasonality, modelling framework and rainfall accumulation period are investigated through an application to Sicily (Italy). In particular, the suitability of different types of regression models used to interpret the relationship between hourly and sub-hourly extreme rainfall and surface temperature is explored.We find overall a sub-CC scaling for most of the island of Sicily. However, the rainfall-temperature scaling relationship is not constant over the temperature range and may be dependent on the season. The different results obtained highlight the importance of modelling choices for analyses in regions characterized by semi-arid climates. More specifically, we observe increasing scaling rates for decreasing rainfall accumulation periods, and significant sensitivity of scaling rates to the selected extreme rainfall quantile. Our novel use of piecewise and locally-weighted scatter plot smoothing regression-based approaches allow the accurate characterization of the temperature dependence of extreme rainfall in Sicily. This identifies a peak-like structure for the drier season, not detected by the simple application of the commonly-used exponential regression based approach.

Use of Regression Models when Performing Fraud Risk Assessment Procedures in the Audit Process

The article provides an overview of current research in the use of regression models when performing assessment procedures of material misstatement risks due to fraud in the financial statement audit. The authors were reviewing regression models predicting deliberate distortion of financial statements, developed by M. Beneish and J. Jones. In addition, they consider the later modifications to these models, applicable in the course of the audit process to estimate the material misstatement risk on the basis of meso-economic, operational, scaling, and other factors affecting the operations of reporting accountants.The specific features, advantages, disadvantages and the use of different types of regression models in the audit process are described. The criteria for comparison are formulated, and the comparative analysis of adapting the best-known features of regression models to the challenges of fraud risk assessment for financial statements in the audit process is carried out. The conclusions about the possibilities of use and a range of different types of regression models when initiating the fraud risk assessment procedures to the financial statements in the audit process are formulated. The limitations, inherent of such models are explained.

Protocol for an economic evaluation of the randomised controlled trial of culprit lesion only PCI versus immediate multivessel PCI in acute myocardial infarction complicated by cardiogenic shock: CULPRIT-SHOCK trial

IntroductionEmergency percutaneous coronary intervention (PCI) of the culprit lesion for patients with acute myocardial infarctions is an accepted practice. A majority of patients present with multivessel disease with additional relevant...

An interactive tool for semi-automatic feature extraction of hyperspectral data

AbstractThe spectral reflectance of the surface provides valuable information about the environment, which can be used to identify objects (e.g. land cover classification) or to estimate quantities of substances (e.g. biomass). We aimed to develop an MS Excel add-in – Hyperspectral Data Analyst (HypDA) – for a multipurpose quantitative analysis of spectral data in VBA programming language. HypDA was designed to calculate spectral indices from spectral data with user defined formulas (in all possible combinations involving a maximum of 4 bands) and to find the best correlations between the quantitative attribute data of the same object. Different types of regression models reveal the relationships, and the best results are saved in a worksheet. Qualitative variables can also be involved in the analysis carried out with separability and hypothesis testing;

A neuro-fuzzy algorithm for improved gas consumption forecasting with economic, environmental and IT/IS indicators

Abstract In energy sector, accurate prediction of long term gas consumption is very important for decision-making and policy process. In addition, conventional approaches may not provide precise results. In this paper, an integrated forecasting algorithm based on Adaptive Neuro Fuzzy Inference System and Computer Simulation (ANFIS-CS) for long term gas consumption has been proposed. Standard input variables include different economic, environmental and IT/IS (number of internet users divided by population in each year) indicators, and the output variable is gas consumption. The concepts of post-processing and pre-processing are considered in the proposed method. At first the best distribution function is identified for each year and then CS is used to create random variables for each year to predict the effects of probabilistic distribution on annual gas consumption. Finally, data is fed into ANFIS model to find the network with the lowest mean absolute percentage error (MAPE). To show the quantitative benefits of the ANFIS-CS, 12 different structures of a well-known class of adaptive neural networks (ANNs), namely Multi-Layer Perceptron (MLP) as well as 10 different types of regression models are developed and the MAPE values of ANN-MLP models and regression models are compared with the MAPE of proposed model. The results of this comparison show the applicability and superiority of the proposed method. This is the first study that presents an integrated intelligent forecasting approach for accurate gas consumption considering the economic, environmental and IT/IS indicators.

Prediction of trihalomethane formation in water distribution systems

Utilities involved in drinking water treatment are faced with a serious challenge in achieving the balance between pathogen destruction by chlorination and trihalomethanes (THMs) minimization below regulatory levels. Mathematical modeling offers an effective tool in cases where the prediction of THM formation is required. The main purpose of this study is to develop an empirical mathematical model that would predict the THM formation in water distribution systems on a laboratory scale. A parallel purpose of the study is to investigate, develop and test procedures for conducting water quality studies related to THM formation. Different types of regression models were investigated using backward elimination for the THM model parameters to select the optimum number of independent variables to be used. The Statistical Package for the Social Sciences backward regression revealed that there are five possible empirical models that predict the amount of total THMs formation as a function of various parameters suc...

Effects of environmental variables on recruitment of anchovy in the Adriatic Sea

The anchovy Engraulis encrasicolus is an important fishery resource in the Adriatic Sea. Fluctuating recruitment of young fish to the stock over time can be related to changes in the environ- ment. The trend of anchovy recruitment in the northern and central Adriatic from 1975 to 2001 was analysed with the aim of identifying possible effects related to 5 environmental factors: surface air temperature, surface atmospheric pressure, quadrant specific wind stresses, Po River runoff and North Atlantic Oscillation (NAO) index. Particular emphasis was placed on 1987, a year of anchovy collapse and fishery crisis. Different types of regression models were applied, both linear and non- linear (simple and multiple), with predictor variables being environmental factors and parental stock abundance. Positive relationships of number of recruits with autumnal SSE and ESE wind stress and both annual and autumnal Po River runoff were found, with a strength comparable to the relationship between recruits and parental stock. Low levels of these environmental factors were observed just before the 1987 collapse, together with a high frequency of occurrence of NE winds and an extreme positive value of the NAO index in the previous autumn (which may have been unfavourable to recruitment in 1987). All 5 environmental factors could be related to increased or reduced food availability for young stages of anchovy in autumn.

An improved strategy for regression of biophysical variables and Landsat ETM+ data

Empirical models are important tools for relating field-measured biophysical variables to remote sensing data. Regression analysis has been a popular empirical method of linking these two types of data to provide continuous estimates for variables such as biomass, percent woody canopy cover, and leaf area index (LAI). Traditional methods of regression are not sufficient when resulting biophysical surfaces derived from remote sensing are subsequently used to drive ecosystem process models. Most regression analyses in remote sensing rely on a single spectral vegetation index (SVI) based on red and near-infrared reflectance from a single date of imagery. There are compelling reasons for utilizing greater spectral dimensionality, and for including SVIs from multiple dates in a regression analysis. Moreover, when including multiple SVIs and/or dates, it is useful to integrate these into a single index for regression modeling. Selection of an appropriate regression model, use of multiple SVIs from multiple dates of imagery as predictor variables, and employment of canonical correlation analysis (CCA) to integrate these multiple indices into a single index represent a significant strategic improvement over existing uses of regression analysis in remote sensing. To demonstrate this improved strategy, we compared three different types of regression models to predict LAI for an agro-ecosystem and live tree canopy cover for a needleleaf evergreen boreal forest: traditional ( Y on X) ordinary least squares (OLS) regression, inverse ( X on Y) OLS regression, and an orthogonal regression method called reduced major axis (RMA). Each model incorporated multiple SVIs from multiple dates and CCA was used to integrate these. For a given dataset, the three regression-modeling approaches produced identical coefficients of determination and intercepts, but different slopes, giving rise to divergent predictive characteristics. The traditional approach yielded the lowest root mean square error (RMSE), but the variance in the predictions was lower than the variance in the observed dataset. The inverse method had the highest RMSE and the variance was inflated relative to the variance of the observed dataset. RMA provided an intermediate set of predictions in terms of the RMSE, and the variance in the observations was preserved in the predictions. These results are predictable from regression theory, but that theory has been essentially ignored within the discipline of remote sensing.