Multilayer Perceptron Method Research Articles

Rapid and non-destructive determination of soluble solid content of crown pear by visible/near-infrared spectroscopy with deep learning regression

The development of rapid and non-destructive prediction technology for fruit quality after harvest could enhance market competitiveness and profitability. The soluble solids content (SSC) is an essential quality index of fruit. This study aims to predict the SSC of crown pear using visible/near-infrared (Vis/NIR) spectroscopy (397–1187 nm) with deep learning model MLP-CNN-TCN. Firstly, the spectral information of crown pears was collected, and Savitzky-Golay (SG) smoothing method and Standard normal variate (SNV) were used to preprocess the spectral data. Secondly, the Multi-layer perceptron (MLP) method was used to reduce the dimension of the preprocessed data, and the reduced data was input into a one-dimensional convolutional neural network (1D-CNN) to extract spectral features. Finally, a Temporal convolutional neural network (TCN) was used to establish a regression model to predict pear SSC. This method was also compared with feedforward neural network (FNN), MLP, 1D-CNN, partial least squares regression (PLSR), and support vector regression (SVR). The MLP-CNN-TCN model obtained better prediction performance, with a prediction determination coefficient (RP2) of 0.956 for SSC. This study demonstrated that the combination of Vis/NIR spectroscopy and MLP-CNN-TCN method could rapidly and non-destructively detect SSC of crown pear, and provide a new regression alternative for the prediction of fruit SSC.

EARLY PREDICTION OF UNEXPECTED LATENT SHOCK IN THE EMERGENCY DEPARTMENT USING VITAL SIGNS.

Objective/Introduction : Sequential vital-sign information and trends in vital signs are useful for predicting changes in patient state. This study aims to predict latent shock by observing sequential changes in patient vital signs. Methods : The dataset for this retrospective study contained a total of 93,194 emergency department (ED) visits from January 1, 2016, and December 31, 2020, and Medical Information Mart for Intensive Care (MIMIC)-IV-ED data. We further divided the data into training and validation datasets by random sampling without replacement at a 7:3 ratio. We carried out external validation with MIMIC-IV-ED. Our prediction model included logistic regression (LR), random forest (RF) classifier, a multilayer perceptron (MLP), and a recurrent neural network (RNN). To analyze the model performance, we used area under the receiver operating characteristic curve (AUROC). Results : Data of 89,250 visits of patients who met prespecified criteria were used to develop a latent-shock prediction model. Data of 142,250 patient visits from MIMIC-IV-ED satisfying the same inclusion criteria were used for external validation of the prediction model. The AUROC values of prediction for latent shock were 0.822, 0.841, 0.852, and 0.830 with RNN, MLP, RF, and LR methods, respectively, at 3 h before latent shock. This is higher than the shock index or adjusted shock index. Conclusion : We developed a latent shock prediction model based on 24 h of vital-sign sequence that changed with time and predicted the results by individual.

Machine learning multilayer perceptron method for building information modeling application in engineering performance prediction

ABSTRACT The engineering design process has fundamentally impacted the life cycle of construction projects and notably, the engineering performance is significantly measured in delivering projects. Previous studies on engineering performance have established the cause–effect relationships between project variables and performance measures. Recently, the building information modeling (BIM) application has reformed how owners execute the engineering, construction, commissioning, and operation in the industry. There has been an increasing focus in finding the benefits of BIM on project performance, however, a minor focus has been given to engineering performance. This paper proposes an artificial neural network (ANN) machine learning multilayer perceptron (MLMP) method and linear regression (LR) that correlates the use of BIM with engineering performance for better construction project assessment. The conclusions reveal a high-level correlation measure between BIM use inputs and engineering performance outputs and further methods for evaluating the engineering performance. Furthermore, we achieved and validated the best prediction by leveraging data from 60 samples using the MLMP and LR models.

An Attention-based Bidirectional LSTM Model for Continuous Cross-Subject Estimation of Knee Joint Angle during Running from sEMG Signals.

Accurate and robust estimation of joint kinematics via surface electromyogram (sEMG) signals provides a human-machine interaction (HMI)-based method that can be used to adequately control rehabilitation robots while performing complex movements, such as running, for motor function restoration in affected individuals. To this end, this paper proposes a deep learning-based model (AM-BiLSTM) that integrates a bidirectional long short-term memory (BiLSTM) network and an attention mechanism (AM) for robust estimation of joint kinematics. The proposed model was appraised using knee joint kinematic and sEMG signals collected from fourteen subjects who performed running at the speed of 2 m/s. The proposed model's generalizability was tested for both within- and cross-subject scenarios and compared with long short-term memory (LSTM) and multi-layer perceptron (MLP) networks in terms of normalized root-mean-square error and correlation coefficient metrics. Based on the statistical tests, the proposed AM-BiLSTM model significantly outperformed the LSTM and MLP methods in both within- and cross-subject scenarios (p<0.05) and achieved state-of-the-art performance.Clinical Relevance- The promising results of this study suggest that the AM-BiLSTM model has the potential for continuous cross-subject estimation of lower limb kinematics during running, which can be used to control sEMG-driven exoskeleton robots oriented towards rehabilitation training.

Experimental and Modeling Evaluation of Impacts of Different Tillage Practices on Fitting Parameters of Kostiakov’s Cumulative Infiltration Empirical Equation

The evaluation and modeling of the water infiltration rate into the soil are important to all aspects of water resources management and the design of irrigation systems for agricultural purposes. However, research focused on experimental studies of infiltration rates in clay soils under different tillage practices remains minimal. Therefore, an empirical prediction model for cumulative water infiltration needs to be created to estimate water depth under different tillage practices. Thus, the present research investigated the impacts of different tillage practices, including plow type (three tillage systems: moldboard, disk, and rotary plows), tillage depth (100 and 200 mm) and four soil compactions levels (0, 1, 3, and 5 tractor wheel passes), on cumulative infiltration behavior in a clay soil under a randomized complete design with three replications. Double-ring infiltration experiments were conducted to collect infiltration data. The research was conducted in three different stages. The first stage was performed through a field test to obtain infiltration data, the second stage involved using a Kostiakov empirical equation (Z = q × tb) for cumulative infiltration to acquire the fitting parameters of “q” and “b”, and in the last stage, we predicted the fitting parameters of “q” and “b” based on soil mean weight diameter, tillage depth, and four soil compaction levels by applying regression data mining approaches in Weka 3.8 software. The results show that the effects of relevant factors on the cumulative water infiltration depth of the soil could be statistically significant (p &lt; 0.05). The Kostiakov model, with an average coefficient of determination of 0.939, had a good fitting effect on the cumulative water infiltration depth process of the investigated soil. The average, lowest, and maximum values of the “q” parameter were 2.7073, 2.2724, and 3.1277 mm/minb, respectively, while for the “b” parameter, they were 0.5523, 0.5424, and 0.5647, respectively. Furthermore, the evaluation of several regression data mining approaches determined that the KStar (K*) data mining approach, with a root mean square error of 0.0228 mm/minb, a mean absolute error of 0.0179 mm/minb, and a correlation coefficient of 0.997, was the most accurate method for fitting parameter “q” using the testing dataset. The most accurate method for fitting the parameter “b” estimation was determined to be the Multilayer Perceptron method, with a root mean square error of 0.0026, a mean absolute error of 0.0013, and a correlation coefficient of 0.962, using the testing dataset. Therefore, this research, which consisted of in situ field observation experiments and infiltration modeling of the infiltration process in a clay soil, provides an essential theoretical basis for improving models of the rate of cumulative infiltration. Moreover, the proposed methodology could be employed for simulation of the fitting parameters “q” and “b” for soil water cumulative infiltration processes, not only for irrigation management purposes under regular crop production conditions, but also for the selection of the most suitable tillage practices to modify the soil during the agriculture season to conserve water and prevent yield declines. The results support the understanding of the infiltration processes in a clay soil and demonstrate that tillage practices could reduce the water infiltration rate into the soil.

Predicting Indian electricity exchange‐traded market prices: SARIMA and MLP approach

AbstractThis research investigates the short‐term (ST) forecasting performance of the daily prices of the Indian exchange‐traded day‐ahead (DAM) market, divided into 13 bid areas, each consisting of states with varied fundamentals. Forecasts are built employing SARIMA (seasonal autoregressive integrated moving average) and MLP (multilayer perceptron) methods. Moreover, the robustness and performance of the model is compared using the lowest error and the Diebold–Mariano (DM) test statistic values. The results indicates that the SARIMA model has high prediction accuracy with error values ranging from 1% to 5% with Southern region having the highest error of 4.53% and Northern having the least error of 1.27%. However, validation by the DM test suggests no statistical significant difference between the two models. The power generators, distribution companies, traders, policymakers, strategists and managers could use the findings for effective power management through proper planning.

Analyzing the Effectiveness of Imbalanced Data Handling Techniques in Predicting Driver Phone Use

Distracted driving leads to a significant number of road crashes worldwide. Smartphone use is one of the most common causes of cognitive distraction among drivers. Available data on drivers’ phone use presents an invaluable opportunity to identify the main factors behind this behavior. Machine learning (ML) techniques are among the most effective techniques for this purpose. However, the potential and usefulness of these techniques are limited, due to the imbalance of available data. The majority class of instances collected is for drivers who do not use their phones, while the minority class is for those who do use their phones. This paper evaluates two main approaches for handling imbalanced datasets on driver phone use. These methods include oversampling and undersampling. The effectiveness of each method was evaluated using six ML techniques: Multilayer Perceptron (MLP), Support Vector Machine (SVM), Naive Bayes (NB), Bayesian Network (BayesNet), J48, and ID3. The proposed methods were also evaluated on three Deep Learning (DL) models: Arch1 (5 hidden layers), Arch2 (10 hidden layers), and Arch3 (15 hidden layers). The data used in this document were collected through a direct observation study to explore a set of human, vehicle, and road surface characteristics. The results showed that all ML methods, as well as DL methods, achieved balanced accuracy values for both classes. ID3, J48, and MLP methods outperformed the rest of the ML methods in all scenarios, with ID3 achieving slightly better accuracy. The DL methods also provided good performances, especially for the undersampling data. The results also showed that the classification methods performed best on the undersampled data. It was concluded that road classification has the highest impact on cell phone use, followed by driver age group, driver gender, vehicle type, and, finally, driver seatbelt usage.

Investigating the effect of surface urban heat island on the trend of temperature changes

Urbanization has led to the emergence of surface urban heat islands (SUHI) and air pollution, necessitating comprehensive investigations into the urbanization process. This study explores the impact of urbanization on temperature trends in northern Iran using remote sensing and neural networks. MODIS images from 2001, 2010, and 2019 were employed to determine climatic features and vegetation levels. SUHI indices and the Urban Thermal Field Variance Index (UTFVI) were utilized to assess ecological comfort and urban heat islands. Markov chain methods were used for forecasting future thermal pollution. The relationship between temperature, air pollutants, spectral indices, and thermal indices was investigated through regression analysis. Furthermore, multilayer perceptron (MLP), radial basis function (RBF), and long short-term memory (LSTM) neural networks were employed for predicting indicator values. The findings reveal an increase in thermal pollution indicated by elevated Land Surface Temperature (LST), SUHI, and UTFVI values, along with a decrease in Normalized Difference Vegetation Index (NDVI) values between 2001 and 2019. The Urbanization Index (UI) values demonstrate a 5% increase in urban areas during the same period. The UTFVI values suggest a decline in ecological comfort due to urbanization and reduced vegetation, with the majority of the region experiencing the poorest thermal comfort in 2019. The SUHI index peaked at 11.72 in 2019, with the highest values observed in spring and the lowest in winter. Correlation analysis highlights a significant relationship between LST and various factors, including SUHI, UTFVI, CH4, SO2, and CO2, underscoring the influence of temperature increase on pollution escalation in the study area. Air pollution assessment indicates the highest concentration of CO gas in the northern and southern parts of the region. Tehran exhibits the highest CO gas levels at 0.051 ppm, while Golestan records the lowest at 0.012 ppm. SO2 levels reach their peak in northern regions (0.0088 ppm) and are lowest in southern regions (0 ppm). The Markov and CA-Markov chain methods predict that by 2050, approximately 50% of the region will experience high temperatures. Neural network results demonstrate that the LSTM method outperforms MLP and RBF methods, with R2 values of 0.99 and 0.98, respectively, indicating higher accuracy in LST prediction during testing and validation stages. Overall, this study highlights the potential of neural networks in predicting thermal pollution and facilitating effective land management strategies in urban areas.

Deep reinforcement learning based spectrum prediction for bursty bands

Spectrum prediction plays an important role for the secondary user (SU) to utilize the shared spectrum resources. However, currently utilized prediction methods are not well applied to spectrum with high burstiness, as parameters of prediction models cannot be adjusted properly. This paper studies the prediction problem of bursty bands. Specifically, we first collect real WiFi transmission data in 2.4GHz Industrial, Scientific, Medical (ISM) band which is considered to have bursty characteristics. Feature analysis of the data indicates that the spectrum occupancy law of the data is time-variant, which suggests that the performance of commonly used single prediction model could be restricted. Considering that the match between diverse spectrum states and multiple prediction models may essentially improve the prediction performance, we then propose a deep-reinforcement learning based multilayer perceptron (DRL-MLP) method to address this matching problem. The state space of the method is composed of feature vectors, and each of the vectors contains multi-dimensional feature values. Meanwhile, the action space consists of several multilayer perceptrons (MLPs) that are trained on the basis of multiple classified data sets. We finally conduct experiments with the collected real data and simulations with generated data to verify the performance of the proposed method. The results demonstrate that the proposed method significantly outperforms the state-of-the-art methods in terms of the prediction accuracy.

Multilayer Perceptron Network Optimization for Chaotic Time Series Modeling.

Chaotic time series are widely present in practice, but due to their characteristics-such as internal randomness, nonlinearity, and long-term unpredictability-it is difficult to achieve high-precision intermediate or long-term predictions. Multi-layer perceptron (MLP) networks are an effective tool for chaotic time series modeling. Focusing on chaotic time series modeling, this paper presents a generalized degree of freedom approximation method of MLP. We then obtain its Akachi information criterion, which is designed as the loss function for training, hence developing an overall framework for chaotic time series analysis, including phase space reconstruction, model training, and model selection. To verify the effectiveness of the proposed method, it is applied to two artificial chaotic time series and two real-world chaotic time series. The numerical results show that the proposed optimized method is effective to obtain the best model from a group of candidates. Moreover, the optimized models perform very well in multi-step prediction tasks.

Investigating patterns of air pollution in metropolises using remote sensing and neural networks during the COVID-19 pandemic

The purpose of this study is to determine the amount of air pollution in Tehran, Isfahan, Semnan, Mashhad, Golestan, and Shiraz during the Corona era and before. For this purpose, Sentinel satellite images were used to investigate the concentration of Methane (CH4), Carbon Monoxide (CO), Carbon Dioxide (CO2), Nitrogen Dioxide (NO2), Ozone (O3), Sulfur Dioxide (SO2), aerosol pollutants in In the era before and during Corona. Furthermore, greenhouse effect-prone areas were determined in this study. In the following, the state of air inversion in the studied area was determined by taking the temperature on the surface of the earth and in the upper atmosphere, as well as the wind speed into account. In this research, the prediction of air temperature for the year 2040 was conducted using the Markov and Cellular Automaton (CA)-Markov methods, considering the impact of air pollution on the air temperature of metropolises. Additionally, the Radial Basis Function (RBF) and Multilayer Perceptron (MLP) methods have been developed to determine the relationship between pollutants, areas prone to air inversions, and temperature values. According to the results, pollution caused by pollutants has decreased in the Corona era. According to the results, there is more pollution in Tehran and Isfahan metropolises. In addition, the results showed that air inversions in Tehran is the highest. Additionally, the results showed a high correlation between temperature and pollution levels (R2 = 0.87). Thermal indices in the studied area indicate that Isfahan and Tehran, with high values of Surface Urban Heat Island (SUHI) and being in the 6th class of thermal comfort (Urban Thermal Field Variance Index (UTFVI)), are affected by thermal pollution. The results showed that parts of southern Tehran province, southern Semnan and northeastern Isfahan will have higher temperatures in 2040 (class 5 and 6). Finally, the results of the neural network method showed that the MLP method with R2 = 0.90 is more accurate than the RBF method in predicting pollution amounts. This study significantly contributes by introducing innovative advancements through the application of RBF and MLP methods to assess air pollution levels during the COVID-19 and pre-pandemic periods, while also investigating the intricate relationships among greenhouse gases, air inversion, air temperature, and pollutant indices within the atmosphere. The utilization of these methods notably enhances the accuracy and reliability of pollution predictions, amplifying the originality and significance of this research.

Application of multilayer perceptron (MLP) method for streamflow forecasting (case study: Upper Citarum River, Indonesia)

Flood forecasting is a critical component of flood early warning. The discharge that occurs is one of the parameters that can be used as a reference for predicting flooding. Various discharge forecasting models based on physically based models or data-driven models have been developed. One of the flood forecasting methods that can be considered for forecasting discharge on watersheds with limited physical data is the Artificial Neural Network (ANN). Furthermore, the ANN method allows the analysis process to be completed in less time and with fewer resources. One of the ANN models employed in this work is the multilayer perceptron (MLP). The MLP model was developed in this study to predict streamflow at the Citarum river, particularly at the Dayeuhkolot hydrological station at 2, 4, 6, 8, 10, 12, and 24 hours ahead. Two data input scenarios were used in the modeling scene. First, input data in the form of station rainfall data and discharge data. The second is regional rainfall and discharge data. Before predicting the discharge in the coming hours, the hyperparameters model is optimized using the GridSearchCV method. The model’s performance is assessed using the RMSE, NSE, and R2 values. The MLP method produced satisfactory results for both scenarios when predicting discharge in less than 4 hours with the NSE and R2 value higher than 0.9. Scenario 2 input data produces a slightly better prediction model than scenario 1. Based on NSE and R2 values, discharge prediction with a prediction time of more than 6 hours produces less accurate results.

Laboratory Data and IBDQ-Effective Predictors for the Non-Invasive Machine-Learning-Based Prediction of Endoscopic Activity in Ulcerative Colitis.

A suitable, non-invasive biomarker for assessing endoscopic disease activity (EDA) in ulcerative colitis (UC) has yet to be identified. Our study aimed to develop a cost-effective and non-invasive machine learning (ML) method that utilizes the cost-free Inflammatory Bowel Disease Questionnaire (IBDQ) score and low-cost biological predictors to estimate EDA. Four random forest (RF) and four multilayer perceptron (MLP) classifiers were proposed. The results show that the inclusion of IBDQ in the list of predictors that were fed to the models improved accuracy and the AUC for both the RF and the MLP algorithms. Moreover, the RF technique performed noticeably better than the MLP method on unseen data (the independent patient cohort). This is the first study to propose the use of IBDQ as a predictor in an ML model to estimate UC EDA. The deployment of this ML model can furnish doctors and patients with valuable insights into EDA, a highly beneficial resource for individuals with UC who need long-term treatment.

Surabaya land cover prediction based on Landsat Satellite using the Multi-layer Perceptron Method

Infrastructure development has occurred very rapidly in Surabaya in the last few decades. There is pressure on the need for land use in line with the increase in population. This study aims to analyze changes in land cover with inter-decade analysis and predict land cover in 2021-2030 in Surabaya using the Multi-Layer Perceptron method. The data used in this research is Landsat satellite which is considered good in representing the actual land cover. A decrease in the amount of vegetation occurs every decade in Surabaya, while the number of buildings is increasing and the bodies of water are relatively the same. The Multi-Layer Perceptron method has a good level of accuracy in predicting land cover in the city of Surabaya. In 2021-2030 it is predicted that Surabaya will still experience an increase in the number of buildings and a decrease in the amount of vegetation.

Accuracy assessment of various supervised machine learning algorithms in litho-facies classification from seismic data in the Penobscot field, Scotian Basin

Litho-facies classification is an essential task in characterizing the complex reservoirs in petroleum exploration and subsequent field development. The lithofacies classification at borehole locations is detailed but lacks in providing larger coverage areas. The acquired 3D seismic data provides global coverage for studying the reservoir facies heterogeneities in the study area. This study applies six supervised machine learning techniques (Random Forest, Support Vector Machine, Artificial Neural Network, Adaptive Boosting, Xtreme Gradient Boosting, and Multilayer Perceptron) to 3D post-stack seismic data to accurately estimate different litho-facies in inter-well regions and compares their performance. Initially, the efficacy of the said models was critically examined via the confusion matrix (accuracy and misclass) and evaluation matrix (precision, recall, F1-score) on the test data. It was found that all the machine learning models performed best in classifying the shale facies (87%–94%) followed by the sand (65%–79%) and carbonate facies (60%–78%) in the Penobscot field, Scotian Basin. On an overall accuracy scale, we found the multilayer perceptron method the best-performing tool, whereas the adaptive boosting method was the least-performing tool in classifying all three litho-facies in the current analysis. While other methods also performed moderately good for the classification of all three litho-facies. The predicted litho-facies using seismic attributes matched well with the log data interpreted facies on the borehole locations. It indicates that the facies estimated in inter-well regions are accurate and reliable. Furthermore, we validated the estimated results with the other seismic attributes to ascertain the accuracy and reliability of the predicted litho-facies between the borehole locations. This study recommends machine learning applications for litho-facies classification to reduce the risk associated with reservoir characterization.

Creation of wildfire susceptibility maps in the Mediterranean Region (Turkey) using convolutional neural networks and multilayer perceptron techniques

Considering the natural disasters that have developed in the world in recent years, it is known that there is an increase in wildfire disasters with the effects of climate change. In this study, wildfire susceptible areas were determined in the provinces of Muğla, Antalya, Mersin, Adana, Osmaniye, and Hatay in the Mediterranean region (Turkey). Within the scope of this purpose, Convolutional Neural Network (CNN) and Multilayer Perceptron (MLP) methods, the most widely used deep learning techniques in the literature in recent years, were preferred to create Wildfire susceptibility models. Seventeen environmental variables were used in the analyses, and these variables were grouped as topographic factors, anthropological and environmental factors, climatic factors, and vegetation factors. In addition, the number of fire inventory data has been balanced with the help of the Synthetic Minority Oversampling Technique (SMOTE) used to increase the model result performance of the scarce inventory data. In the maps obtained by CNN and MLP methods, 17% and 28% of the study area were determined as high and very high susceptible areas, respectively. The results demonstrated that the CNN model had superior performance in Wildfire susceptibility assessment with accuracy (%85.8), precision (%98.7), sensitivity (%85.5), F- Score (%91.6), and ROC curve (%78.6). This model was followed by the MLP model with slightly lower accuracy values, which indicates that the CNN models can reach considerably better prediction capability than the MLP models. Finally, the wildfire susceptibility maps produced by deep learning methods could aid decision-makers and government organizations in the Mediterranean region in preventing future natural disasters.

Forecasting of Average Monthly Flow for Two Stations at Khabour River Using Arima And Ann Models

In this study, autoregressive integrated moving average (ARIMA) and artificial neural network (ANN) models were applied to predict the average monthly flow time series of two stations, named Begova and Chemcermo, for the Khabour River. The analysis of the time series was performed using several criteria and tests. The autocorrelation function (ACF) and partial autocorrelation function (PACF) were applied to check the accuracy of the ARIMA model. The Akaike (AIC) and Bayesian (BIC) equations were performed to determine the optimum model for prediction, which depended on the lowest AIC and BIC values. Applied test results show that the models of order ARIMA (0,0,0)(3,0,0)12 and ARIMA((0,0,5)(5,1,4)12 have higher acceptance compared to the other models for predicting the average monthly flow for Begova and Chemecermo stations, respectively. An ANN model of type multilayer perceptron method (ANN-MLP) was used for predicting the average monthly flow of Begova and Chemecermo stations, where the best models found are MLP (5,3,1) and MLP (9,7,1), respectively. Different statistical tests were applied and showed that the efficiency of the ANN model was better than the ARIMA model, with deterministic coefficients of 0.914 and 0.876 compared to 0.854, and 0.852 for the ARIMA for the monthly time series of Begova and Chemecermo stations, respectively

Forecasting of Average Monthly Flow for Two Stations at Khabour River Using Arima And Ann Models

In this study, autoregressive integrated moving average (ARIMA) and artificial neural network (ANN) models were applied to predict the average monthly flow time series of two stations, named Begova and Chemcermo, for the Khabour River. The analysis of the time series was performed using several criteria and tests. The autocorrelation function (ACF) and partial autocorrelation function (PACF) were applied to check the accuracy of the ARIMA model. The Akaike (AIC) and Bayesian (BIC) equations were performed to determine the optimum model for prediction, which depended on the lowest AIC and BIC values. Applied test results show that the models of order ARIMA (0,0,0)(3,0,0)12 and ARIMA((0,0,5)(5,1,4)12 have higher acceptance compared to the other models for predicting the average monthly flow for Begova and Chemecermo stations, respectively. An ANN model of type multilayer perceptron method (ANN-MLP) was used for predicting the average monthly flow of Begova and Chemecermo stations, where the best models found are MLP (5,3,1) and MLP (9,7,1), respectively. Different statistical tests were applied and showed that the efficiency of the ANN model was better than the ARIMA model, with deterministic coefficients of 0.914 and 0.876 compared to 0.854, and 0.852 for the ARIMA for the monthly time series of Begova and Chemecermo stations, respectively

Computational simulation of organic pollutant removal from wastewater by different porous boron nitride (PBN) nanoclusters

A computational modeling framework was developed to evaluate the adsorption behavior of boron nitride (BN) for separation of a solute from solution. Adsorption of an organic dye onto the surface of different porous BNs at two different temperatures was computed using different machine learning models. The amount of dye concentration changes in the solution based on the time parameter for different materials has been modeled and validated in this research. An attempt was made to construct the kinetics of adsorption using the developed models rather than the conventional kinetic models. A decision tree (DT), Multilayer Perceptron (MLP), and Linear Regression boosted with Adaboost (ADA-LR) are selected models that are optimized by their hyper-parameters using a grid-search. Models were developed separately for low and high temperatures for three different materials with different chemistry and structures. Using the MSE parameter, DT has an error of 1.44E-02 for low temperatures and 1.48E-02 for high temperatures. It is 1.01E-03 for low and 7.44E-03 for high temperatures using MLP method of simulation. However, in terms of MSE, the best model is ADA-LR, with 4.60E-03 for low temperature and 1.81E-03 for high temperature. With other metrics also, ADA-LR is the best model, so it is selected as the main model of this research. The R2-score of the final model is 0.940 for low and 0.973 for high temperature values.

Sentiment Analysis and Twitter Social Media Visualization Regarding the Omnibus Law Draft

This study will classify Twitter users' positive and negative opinions about the omnibus method using a frequency-inverse document frequency algorithm and a multi-layer perceptron method. The sentiment analysis process involves several stages, including B. Collecting and preprocessing data, calculating term weights using inverse term frequencies and document frequencies, and classifying data using multi-layer perceptrons. Additionally, the study visually represents Twitter's sentiment analysis results on the omnibus method. These visualizations include word cloud, top accounts, tweet frequency, hashtags, and sentiment. Three scenarios were considered to perform the classification experiments. Scenario 1 used 700 training data, scenario 2 used 800, and Scenario 3 used 900 training data. The findings show that the Term Frequency Inverse Document Frequency algorithm and the multi-layer perceptron method are adequate for sentiment analysis, with Scenario 3 yielding the highest accuracy rate of 88%.