The object of this study was to compare predictive accuracy of some individual modeling methods versus ensemble modeling approach in estimating the spatial potential distribution and identifying ecological requirements of Artemisia sieberi in desert rangelands of Yazd province, Central Iran. For this purpose, the species presence data were collected using the random systematic sampling method in 2019. Individual modeling of the species distribution was performed using Random Forest, Classification and Regression Tree and Generalized Additive Model after preparing environmental variable maps using GIS and geostatistics. Predictive performance of individual models was evaluated using Area Under Curve and Root Mean Square Error statistics. Furthermore, the Ensemble model was used based on the weighted average AUC. The appropriate threshold limit value was calculated based on True Skill Statistic for conversion of continuous maps to binary ones of habitat suitability. Comparison of the performance of individual models showed that the RF model had a more accurate prediction compared to the other models (AUC= 0.971 and RMSE= 0.256). Evaluation of the models implemented using threshold-dependent metrics such as Sensitivity, Specificity, and Kappa index also confirmed this finding. The overall comparison of the results from the three models versus the Ensemble model also indicates the high performance of this model compared to the individual models. Based on Ensemble model results, 45.38% of the study area had a high suitability for the establishment of A. sieberi. Based on the analysis of the importance of variables in the RF model, elevation (42%), Clay (40.02%) and pH (38.97%) in 0-30 cm soil depth had the highest effect on the presence of species. In general, Ensemble modeling can reduce the uncertainty and provide more reliable results by combining the results of the different algorithms of individual modeling.

Failure of earth fill dams is a great challenge in Civil Engineering, in which one of the main causes is uncontrolled seepage through the core and foundation of the dam. Thus seepage analysis is one of the most important complications in design, construction and operation of this type of dams; in this way, inspecting the piezometric heads is the first step in seepage analysis. In the following paper, Sattarkhan earth fill dam piezometric heads have been analyzed via Artificial Intelligence (AI) models and a classic black box model, based on two scenarios. Each scenario has different input combinations for modeling of various conditions. To continue ensemble models have been formed via outputs of the single black box models to improve modeling performance. Three methods of model ensemble were considered, including simple linear averaging model, weighted linear averaging model and non-linear neural ensemble model. Results show that employing model ensemble and in particular non-linear ensemble by neural network, improve the modeling accuracy up to 10%. Moreover, by comparison the both scenarios, it is concluded that in case of a failure of a piezometer, employing scenario 2 can be an effective way.

In this study, black box artificial intelligence models (AI) including feed-forward neural network (FFNN), support vector regression (SVR), and adaptive neuro-fuzzy inference system (ANFIS) were used to predict effluent biological oxygen demand (BODeff) and chemical oxygen demand (CODeff) of Tabriz wastewater treatment plant (WWTP) using the daily data collected from 2016 to 2018. In addition, the autoregressive integrated moving average (ARIMA) linear model was used to predict BODeff and CODeff parameters to compare the linear and non-linear models' abilities in complex processes prediction. To improve the prediction of BODeff and CODeff parameters, the data post-processing ensemble method was also used. The input data set included daily influent BOD, COD, total suspended solids (TSS), pH at the current time (t), and BODeff and CODeff at the previous time (t-1) and the output data included BODeff and CODeff at t. The results of the single models indicated that the SVR model provides better results than the other single models. In ensemble modeling, simple and weighted linear averaging, and neural network ensemble methods were applied to enhance the performance of the single AI models. The results indicated that using ensemble models could increase the prediction accuracy up to 15% at the verification phase.

This study evaluates the performance of an ensemble framework using the Weather Research and Forecasting (WRF) model for probabilistic seasonal precipitation forecasts. In this study, two types of data were used: a) The meteorological initial and boundary conditions come from the National Centers for Environmental Prediction (NCEP) Climate Forecast System Version 2 (CFSv2) data. b) Precipitation data from the Global Precipitation Climatology Centre (GPCC) dataset used as observational data over Iran. The ensemble model was designed based on a one-way double-nested (60-parent domain and 20-nested Km resolutions) modeling system using Weather Research and Forecasting (WRF) version 4.2 customized over Iran to downscale the second version of the NCEP Climate Forecast System (CFSv2). The results showed that precipitation forecast at seasonal time scale in Iran has high uncertainty. Although probabilistic forecasts can increase the efficacy of seasonal forecasts more than deterministic, the uncertainty of these forecasts is still high. Additionally, the downscaling of the CFS.v2 model by WRF and using multiple initial conditions and model physics can increase the accuracy of seasonal forecasts. The spatial distribution of the forecast accuracy of the ensemble model is dependent on the spatial distribution of precipitation over Iran. Another factor that affects the model's accuracy is the forecast lead time dependent especially at 2-month and 3-month forecast lead times. The results showed that the model has high uncertainty in the east and southeast of Iran. The implementation of this model for an operational period showed that although the model can forecast the spatial variation of rainfall over Iran up to a three-month lead time, probabilistic forecasting cannot significantly reduce the uncertainty of the model in a seasonal time scale. The ensemble model tends to overestimate precipitation in the third lead time.

Species distribution model is important to protect and to manage the roe deer as an elusive and national protected herbivore species in Iran. In this study habitat modeling was carried out via Biomod2 package in R software environment using 91 presence points of the spices as well as environmental and human factors. Ensemble model of habitat suitability was prepared based on six species distribution model algorithms for a study area of 74000 km2. AUC for ensemble model was 0.97 which showed an excellent performance. Suitable habitat for roe deer covered an area about 15 percent of the studied area. Also, the results showed that altitude (29%) and land-use (29%) then, maximum temperature in the warmest month and slope variables were the most influential factors on species distribution. Also, the results indicate that anthropogenic factors were not much effective on the species presence. Roe deer prefers the denser forests since this habitat can reduce its visibility and provide sufficient cover and food for the species.

This study investigates the capability of frequency ratio and an ensemble method of frequency ratio with Shannon’s entropy to produce a reliable map of wildfire susceptibility for Chaharmahal and Bakhtiari province, Iran. At first, the fire locations were identified in the study area from historical archives and field surveys. Ninety two cases (70%) out of 132 detected fire locations were randomly selected for modeling, and the remaining 40 (30 %) cases were used for the validation. Thirteen fire conditioning factors representing topography, climate, and human activities of the study area were extracted from the spatial database. Using the frequency ratio and the ensemble model, the relationship between the conditioning factors and fire locations were explored. The results were then used to produce distribution maps of wildfire hazard. The verification analysis using Receiver Operating Characteristic (ROC) curves and the Areas Under the Curve (AUC) revealed that the ensemble model with the capability of computing the weights of factors and their categories is more efficient than frequency ratio. The success and prediction rates for the frequency ratio and ensemble model were found to be 79.2 and 75.72%, and 85.16 and 82.92%, respectively. Further, the results suggest that more than one-third of the study area falls into the high and very high hazard classes, and the conditioning factors of land use, soil types, and distance from roads play major roles in fire occurrence and distribution in the study area.

The Egyptian vulture, Neophron percnopterus, as a globally endangered species, is significantly affected by conflicts with human, leading to the decline of this species population. Availability of food resources is known as one of the most important factors affecting the distribution of the Egyptian vulture. In this study, we employed an ensemble modelling approach to predict the  suitability of the  foraging habitat for the Egyptian vulture based on seven distribution modelling algorithms. According to our results, suitable foraging habitats for this bird accounted for about 4.19% (1021.40 km2) of the study area, of which only approximately 31.92% (326 km2) was covered by the conservation areas. Livestock density (31.82%), the presence of wild ungulates (15.36%), distance to dump sites (14.77%), landcover (12.46%) and elevation (11.85%) made the highest contribution to the ensemble model of this species. Our results indicated that Egyptian vultures mainly rely on dump sites, nomadic areas and livestock farming sites, thereby reflecting the vulture's tendency to livestock carcasses. Thus, it is important to develop effective strategies for proper monitoring and managing human activities in these areas.

Ensemble modelling is expanding in several areas of engineering, especially different aspects of water engineering. Accurate estimation of saffron water requirement (SWR), an essential strategic production of the agriculture sector, is a crucial and influencing act in local water planning of this region. Hence, this study aimed to check the applicability of ensemble modelling in enhancing SWR at Birjand, Southern Khorasan, Iran. The actual water requirement of saffron was recorded in the field lysimetric laboratory at the University of Birjand. The simulation of water requirement was conducted utilizing Decision Tree Regression (DTR) with input climate features. Additionally, Boosting and Bagging methods were employed to establish and enhance the ensemble process of soil water requirement (SWR) simulations. To track the effectiveness of any method, some comparative tests were designed, such as statistical criteria (RMSE and MAE) detection, Violin plot analysis, over/underestimation, times series comparison, and error improvement test. Results indicated that although the acceptable performance of DTR in simulating SWR, the probable improvement was potentially felt. Derived results confirmed that supervised ensemble modelling (Boosting) could enhance the accuracy of DTR by more than 30 percent (reducing absolute error from 36 mm to 23.65 mm), resulting in declining RMSE from 0.44 mm to 0.07 mm. Further, different experiment outcomes revealed that the Boosting algorithm quality is more appealing than DTR and Bagging outputs.

The nature and importance of user’s comments in various social media systems play an important role in creating or changing people's perceptions of certain topics or popularizing them. It has now an important place in various fields, including education, sales, prediction, and so on. In this paper, Facebook social network has been considered as a case study. The purpose of this study is to predict the volume of Facebook users' comments on the published content called post. Therefore, the existing problem is classified as a regression problem. In the method presented in this paper, three regression models called elastic network, M5P model, and radial basis function regression model are combined and an ensemble model is made to predict the volume of comments. In order to combine these base models, a strategy called stack generalization is used, based on which the output of the base models is provided to a linear regression model as new features. This linear regression model combines the outputs of the 3 base models and determines the final output of the system. To evaluate the performance of the proposed model, a database of the UCI dataset, which has 5 training sets and 10 test sets, has been used. Each test set in this database has 100 records. In the present study, the efficiency of the base models and the proposed ensemble model is evaluated on all these sets. Finally, it is concluded that the use of the ensemble model can reduce the average correlation coefficient (as one of the evaluation criteria of the model) to 74.4 ± 16.4, which is an acceptable result.

In recent decades, the ability of meteorological forecasts due to urban development and climate change has become an important issue in human societies. Forecasting these variables in addition to informing different segments of society, plays an effective role in better decision making and planning in different areas, such as water resources management. Nowadays in this field, the use of numerical forecasting models is one of the most common approaches. In this study, the output of seven forecast models from the North American Multi Model Ensemble (NMME) project is used to evaluate the forecast of precipitation in Karkheh Dam basin during the 31 years period (1985-2015). For this purpose, the outputs of precipitation forecast models for 1-3 month lead times were downscaled using Multiple Linear Regression (MLR) method. Then the possibility of improving forecast skill of them has been evaluated using an ensemble model approach. The models were assessed using both continuous and probabilistic methods using Pearson correlation coefficient (ρ) and normalized root mean square error (NRMSE) and reliability diagram respectively. The results indicate that none of the forecast models performed well in all lead times alone and their individual use did not show good performance; while the Multi Model Ensemble (MME) generally shows better performance than the individual models. The results of this study, demonstrate the importance of using an ensemble model based on the outputs of several models for improving the long term precipitation forecast skill.