Tree Modeling Technique Research Articles

Modelling water quality parameters using model tree, random forest, and non-linear regression for Mula-Mutha River, Pune, India.

Evaluation of vital water-quality indicators, especially biological and chemical demand of oxygen (BOD and COD), is important for environmental factors, human health, and agricultural output. In the recent past, data-driven techniques (DDT) offer the ability to automate water quality assessment with more reliable and rapid evaluation. The present study thus aims to utilize various DDTs: random forest (RF), model tree (MT), and non-linear-regression (NLR) to predict vital water quality indicators such as BOD and COD for the three stretches of Mula-Mutha River, Pune, India. Since the river has three stretches: Mutha, Mula, and Mula-Mutha respectively, BOD-COD models have been developed separately for each using MT, RF, and NLR. Data analysis using a violin diagram is done to understand the data characteristics. Further, the models developed were developed using the appropriate input parameters for predicting BOD and COD. Error measures including coefficient of correlation (R), mean absolute error (MAE), and root mean square error (RMSE) were used to evaluate the constructed models. The Taylor diagram, scatter plot, and hydrograph were also used for visual performance analysis. The findings suggest that the MT and RF techniques exhibit a stronger connection between the actual and anticipated levels of BOD and COD, with NLR following closely behind. Practical acceptance of these approaches is increased by RF in the form of trees, MT with an output in the form of a sequence of equations, and NLR with a single equation. These findings help us gain insight into DDT's water quality assessment model, which will also help future researchers and water quality professionals make decisions.

Prevalence of Factors Related to Depressive Symptoms Among Married Individuals.

Background There are multiple studies that indicate that the psychological well-being of a couple and their life satisfaction dependon the family and society. Various factors such as family, family values, marriage style, married life, and education have a great impact on people's lives both directly and indirectly.It is important to understand the effects of these factors on married individuals'lives that lead to depression so that appropriate measures can be taken for its prevention. Objectives This research aims to find the relationship of depressive symptoms among married individuals with various factors such as their marriage style, education, and having children. Materials and methods The study included 433 married individuals from Istanbul who met the criteria for depression. The early identification and prediction of depression in married individuals have been demonstrated to benefit significantly from machine learning techniques. In this study, we used decision tree (DT) and random forest (RF) predictive modeling techniques to create a model to predict the occurrence of depression among married individuals. Results The accuracy of the DT approach was found to be 80%, and the RF approach was 60%. Our results showed that as compared to conventional statistical methods, machine learning models performed better for classifying couples. Conclusion Future support systems that employ a range of data sources to identify individuals who are extremely susceptible to developing depression among married people may be developed using these effective models.

Tree Based Approaches for Predicting Concrete Carbonation Coefficient

Carbonation is one of the critical durability issues in reinforced concrete structures in terms of their structural integrity and safety and may cause the fatal deterioration and corrosion of steel reinforcement if ignored. Many researchers have performed a considerable number of studies to predict the carbonation of concrete structures. However, it is still challenging to predict the carbonation depth or carbonation coefficient, as they depend on various factors. Therefore, creating a model that can learn from available data using Data Driven Techniques (DDT) is a step forward in this research field. This study provides new approaches to predict the carbonation coefficient of concrete through Model Tree (MT), Random Forest (RF) and Multi-Gene Genetic Programming (MGGP) approaches. With 827 case studies, the predicted models can be seen as a function of a set of conditioning factors, which are statistically significant in explaining the carbonation mechanism. The results obtained through MT, RF and MGGP were compared with those obtained through Multiple Linear Regression (MLR), Artificial Neural Networks (ANNs) and Genetic Programming (which were previously developed). The results reveal that the MT, RF and MGGP perform better than the previous models. Moreover, the MT technique displays its output in terms of series of equations, RF as multiple trees and MGGP in form of a single equation, which are more user-friendly and applicable in practice.

Evaluation and Validation of Estimated Sediment Yield and Transport Model Developed with Model Tree Technique

This study evaluated the applicability of existing sediment yield and transport estimation models developed using data mining classification and prediction techniques and validated them. Field surveys were conducted by using an acoustic Doppler current profiler and laser in situ scattering and transmission at measuring points in the main stream of the Nakdong River located where the tributaries of the Geumho, Hwang, and Nam Rivers join. Surveys yielded estimations of water velocity, discharge, and suspended sediment concentrations were measured. In contrast with models based on the general watershed characteristics factors, some models based on hydraulic explanatory flow variables demonstrated an excellent predictability. This is because the selected submodels for validation, which provided excellent prediction results, were based on a large number of calibration data. It indicates that a sufficient number of reliable data is required in developing a sediment yield estimation model using data mining. For practical applications of data mining to extant sediment yield estimation models, comprehensive considerations are required, including the purpose and background of model development, and data range. Furthermore, the existing models should be periodically updated with the consideration of temporal and spatial lumping problems.

Design of Open Pit Mines using 3D Model in Two-element Deposits under Price Uncertainty

When it comes to evaluating mining projects, uncertainty plays a significant role, particularly in the analysis of mining economic characteristics, which makes the assessment of a mining project erroneous and untrustworthy. The volatility of mineral prices is a major cause of economic ambiguity and concern. Economic uncertainty has extensively been examined in mining production project planning, but the majority of the study has focused on single-element deposits, with little emphasis devoted to the significance of pricing uncertainty in two-element deposits. Using a three-dimensional tree model, this study investigates how design could be affected by the pricing uncertainty of two different elements. In this model, not only annual volatility but also monthly volatility were considered due to momentary changes in the price of several elements. To authenticate the proposed model, a numerical example was resolved using discounted cash flow, binomial tree, pyramid tree, and three-dimensional modeling techniques. The results of each approach were compared to those of real-world data. Following the findings of the current investigation, it can be concluded that the values derived from the suggested model (a net present value of $ 324.2 thousand) are more precise than the values acquired from other approaches, and that they are just 8% out of step with reality. Other methods, on the other hand, come up with results that are at least 17% and at most 39% different from those that come from real data.

Soil Infiltration Rate Impact on Water Quality Modeled Using Random Forest Regression

In this paper, Infiltration rate of the soil is investigated by using predictive models of Random forest regression and their performance were compared with Artificial neural network (ANN) and M5P model tree techniques. We utilized 132 field measurements comprising this dataset. 88 models were trained using observations, while the remaining 44 were used to validate it. The cumulative time (Tf), the impurity type (It), the impurity concentration (Ci), and the moisture content (Wc) were utilized as input variables, and the rate of infiltration was employed as the output. To evaluate the efficiency of the two modeling methodologies, correlation coefficients we estimated root mean square error (RMSE), mean absolute error (MAE), relative absolute error (RAE), and root relative square error are all terms that may be used to describe errors (RRSE). The random forest regression approach outperforms the other two models when compared to evolutionary data (ANN and M5P model tree). Using a random forest as a model, regression can properly estimate the infiltration rate within a 25% error range. According to the results of the sensitivity research, cumulative time plays an important influence in determining the soil's penetration rate.

Prediction of bearing capacity of H shaped skirted footings on sand using soft computing techniques

Purpose: The present study aims to apply soft computing techniques, Artificial Neural Network (ANN) and M5P model tree, to predict the ultimate bearing capacity of the H plan shaped skirted footing on the sand Design/methodology/approach: A total of 162 laboratory test data for the regular plan shaped (square, circular, rectangular, and strip (up to L/B = 2.5) skirted footing were collected from the literature to develop the soft computing-based models. These models were later modified for the H Plan shaped skirted footing with the introduction of the multiplication factor. The input variables chosen for the regular plan shaped footings were skirt depth to width of the footing ratio (Ds/B), friction angle of the sand (􀉭), the ratio of the interface friction angle-to-friction angle of sand (δ/􀉭), and length-to-width (L/B) ratio of the footing. The output is the bearing capacity ratio (BCR, a ratio of the bearing capacity of the skirted footing to the bearing capacity of un-skirted footing). Findings: Sensitivity analysis was carried out to see the impact of the individual variable on the BCR). The sensitivity results reveal that the skirt depth to width of the footing ratio is the primary variable affecting the BCR. Finally, the performance of the developed soft computing models was assessed using six statistical parameters. The results from the statistical parameters reveal that model developed using ANN was performing superior to the one prepared using M5P model tree technique for the prediction of the ultimate bearing capacity of H plan shaped skirted footing on sand. Research limitations/implications: The model equations are developed with experimental laboratory data. Hence, these equations need further improvement by using field data. However, until now there no field data have been available to include in the present data set. Practical implications: These proposed model equations can be used to predict the bearing capacity of the H-shaped footing with the help of Ds/B, 􀉭, δ/􀉭 and L/B without performing the laboratory experiments. Originality/value: There is no such model equation that was developed so far for the H-shaped skirted footings. Hence, an attempt was made in this article to predict the bearing capacity of the H-shaped footing by using available experimental data with the help of soft computing techniques.

Estimation of the compressive strength of concretes containing ground granulated blast furnace slag using hybridized multi-objective ANN and salp swarm algorithm

The use of supplementary cementitious materials such as ground granulated blast furnace slag (GGBFS) in concrete mixtures provides many technical and economic benefits. The use of GGBFS as a partial replacement for cement in concrete mixtures can also decrease the energy consumption and reduce the greenhouse gas emissions. Developing an accurate model for estimating the compressive strength of the concretes containing GGBFS is a necessity since the value of the compressive strength is a required parameter in various design codes. Besides, a predictive model for the compressive strength instead of direct laboratory-based measurements can save in energy, cost, and time. Artificial neural network (ANN) algorithm was used in this research to develop a model for the estimation of the compressive strength of concretes containing GGBFS. To optimize the error and complexity of the developed ANN models, a multi-objective slap swarm algorithm (MOSSA), as a multi-objective optimization method, was proposed. The M5P model tree algorithm, as one of the most used classification techniques to solve engineering problems, was also used to develop predictive models of compressive strength. The efficiency of the proposed model developed based on the ANN algorithm was compared with that of the model developed based on the M5P model tree technique using various error measures. The findings from this research indicate that the M5P model tree and the proposed ANN model can successfully provide predictive tools for estimating the compressive strength of concretes containing GGBFS with 12.05% and 7.25% mean absolute percentage error (MAPE), respectively. These values indicate that the proposed model based on ANN algorithm has superior efficiency compared to the one developed using M5P model tree.

New formulation of reduction factors for pile groups under lateral loading through Model Tree technique

ABSTRACT Under lateral loading, the p-y curves of a pile group differ from those of a single pile due to group effects. Hence, p-y curves of single piles are modified by reduction factors for either a row or a group of piles. In order to improve the available approaches for estimation of reduction factors, a database was compiled from the experimental test results reported in the literature. Accordingly, a robust machine learning technique i.e. the M5’ Model Tree method was used to propose new formulas for prediction of row and group reduction factors for pile groups. The new relationships are simple, physically sound and accurate based on different statistical measures. These also account for additional parameters of soil properties that affect the reduction factors. The new models compare favorably with recommendations of design guidelines. Furthermore, considering different levels of risks, appropriate safety factors that can be useful for engineers are presented.

Intermittent Reservoir Daily Inflow Prediction Using Stochastic and Model Tree Techniques

The present study depicts the applicability of model tree (MT) technique to a large data set having large number of zero values. It is also aimed to develop a model that results in simple equations as that of stochastic models. The performance of MT is compared with conventional autoregressive integrated moving average (ARIMA) models. Forty-nine years of daily inflow data from Koyna Reservoir located in Maharashtra, India, are used for developing and testing the models. In this case study of developed MT models, the number of inputs is selected by trial and error and is varied from one lag to eight lags. Numerous MT models were developed by considering the model formulations of pruning and smoothing, whereas in ARIMA model, the number of inputs required for proper modeling is selected from autocorrelation function and partial autocorrelation function plots as well as through trial-and-error procedure. The performances of the developed models were evaluated using various statistical measures. On comparing the daily time step MT and ARIMA models, it is found that un-pruned and un-smoothed MT models performed better than ARIMA models. Even though the number of leaves (local linear equations with nonlinear way of finding them) is slightly larger, the low and peak values of the reservoir inflow are predicted better by MT model. From the results, it is concluded that for better modeling and to have a set of linear applicable equations for smaller time step reservoir inflow, MT technique can be a better choice than ARIMA model.

Quantitative Evaluation of Cyber-Attacks on a Hypothetical School Computer Network

This paper presents the attack tree modeling technique of quantifying cyber-attacks on a hypothetical school network system. Attack trees are constructed by decomposing the path in the network system where attacks are plausible. Considered for the network system are two possible network attack paths. One network path represents an attack through the Internet, and the other represents an attack through the Wireless Access Points (WAPs) in the school network. The probabilities of success of the events, that is, 1) the attack payoff, and 2) the commitment of the attacker to infiltrate the network are estimated for the leaf nodes. These are used to calculate the Returns on Attacks (ROAs) at the Root Nodes. For Phase I, the “As Is” network, the ROA values for both attack paths, are higher than 7 (8.00 and 9.35 respectively), which are high values and unacceptable operationally. In Phase II, countermeasures are implemented, and the two attack trees reevaluated. The probabilities of success of the events, the attack payoff and the commitment of the attacker are then re-estimated. Also, the Returns on Attacks (ROAs) for the Root Nodes are re-assessed after executing the countermeasures. For one attack tree, the ROA value of the Root Node was reduced to 4.83 from 8.0, while, for the other attack tree, the ROA value of the Root Node changed to 3.30 from 9.35. ROA values of 4.83 and 3.30 are acceptable as they fall within the medium value range. The efficacy of this method whereby, attack trees are deployed to mitigate computer network risks, as well as using it to assess the vulnerability of computer networks is quantitatively substantiated.

Predictors of intentional intoxication using decision tree modeling analysis: a retrospective study.

ObjectiveThe suicide rate in South Korea is very high and is expected to increase in coming years. Intoxication is the most common suicide attempt method as well as one of the common reason for presenting to an emergency medical center. We used decision tree modeling analysis to identify predictors of risk for suicide by intentional intoxication.MethodsA single-center, retrospective study was conducted at our hospital using a 4-year registry of the institute from January 1, 2013 to December 31, 2016. Demographic factors, such as sex, age, intentionality, therapeutic adherence, alcohol consumption, smoking status, physical disease, cancer, psychiatric disease, and toxicological factors, such as type of intoxicant and poisoning severity score were collected. Candidate risk factors based on the decision tree were used to select variables for multiple logistic regression analysis.ResultsIn total, 4,023 patients with intoxication were enrolled as study participants, with 2,247 (55.9%) identified as cases of intentional intoxication. Reported annual percentages of intentional intoxication among patients were 628/937 (67.0%), 608/1,082 (56.2%), 536/1,017 (52.7), 475/987 (48.1%) from 2013 to 2016. Significant predictors identified based on decision tree analysis were alcohol consumption, old age, psychiatric disease, smoking, and male sex; those identified based on multiple regression analysis were alcohol consumption, smoking, male sex, psychiatric disease, old age, poor therapeutic adherence, and physical disease.ConclusionWe identified important predictors of suicide risk by intentional intoxication. A specific and realistic approach to analysis using the decision tree modeling technique is an effective method to determine those groups at risk of suicide by intentional intoxication.

Housing demand forecast based on income section using model tree technique

Background/Objectives: Mankiw and Weil modified model, which is mainly used in the field of housing demand, has the problem that the added variable has no linear relationship with the age-specific house demand.Methods/Statistical analysis: In this research, we tried to complement the existing model by proposing aM-W modified model utilizing the Model tree technique. In addition, many poor people need another analysis that understands the characteristics to live in abnormal houses. And, we tried to avoid this problem by reflecting income section. We compare the performance with existing models using the 2005 and 2010 Population and Housing Cencus data.Findings: First, the error rate of the M - W modified model is greatly affected by the extreme poverty class and the low income class. Second, overall the performance of the model tree dominates, the performance has further improved to produce more of the nodes.In the middle class in which five nodes were created, the error rate decreased by 89%, and the correlation coefficient increased by 0.2566 with 0.0490.Third, it is more accurate to use the "total of income section predicted values" rather than the existing "entire section predicted value". Fourth, in order to express an accurate section error, we propose to judge "not the total of income section errors" but "total absolute value of income section error".Improvements/Applications: In this research, there is a limitation that generalization of results is inappropriate. For further research, it is considered appropriate to apply the Random forest method to generalize the results.

Modelling of impact of water quality on infiltration rate of soil by random forest regression

In this paper, Infiltration rate of the soil is investigated by using predictive models of Random forest regression and their performance were compared with Artificial neural network (ANN) and M5P model tree techniques. A dataset consists of 132 field measurements were used. Out of 132 observations randomly selected 88 observations were used for training, whereas remaining 44 were used for testing the model. Input variables consist of cumulative time (Tf), type of impurities (It), concentration of impurities (Ci), and moisture content (Wc) whereas the infiltration rate was considered as output. Correlation coefficient (CC), root mean square error (RMSE), mean absolute error (MAE), relative absolute error (RAE) and root relative square error (RRSE) were considered to compare the performance the both modelling approaches. The result of evolution suggests that Random forest regression approach works well than the other two models (ANN and M5P model tree). The estimated value of infiltration rate using Random forest regression lies within ±25% error lines. Sensitivity analysis suggests that cumulative time is an important parameter for predicting the infiltration rate of the soil.

Predicting cyanobacterial abundance, microcystin, and geosmin in a eutrophic drinking-water reservoir using a 14-year dataset

ABSTRACTHarris TD, Graham JL. 2017. Predicting cyanobacterial abundance, microcystin, and geosmin in a eutrophic drinking-water reservoir using a 14-year dataset. Lake Reserve Manage. 33:32-48.Cyanobacterial blooms degrade water quality in drinking water supply reservoirs by producing toxic and taste-and-odor causing secondary metabolites, which ultimately cause public health concerns and lead to increased treatment costs for water utilities. There have been numerous attempts to create models that predict cyanobacteria and their secondary metabolites, most using linear models; however, linear models are limited by assumptions about the data and have had limited success as predictive tools. Thus, lake and reservoir managers need improved modeling techniques that can accurately predict large bloom events that have the highest impact on recreational activities and drinking-water treatment processes. In this study, we compared 12 unique linear and nonlinear regression modeling techniques to predict cyanobacterial abundance and the cyanobacterial secondary metabolites microcystin and geosmin using 14 years of physiochemical water quality data collected from Cheney Reservoir, Kansas. Support vector machine (SVM), random forest (RF), boosted tree (BT), and Cubist modeling techniques were the most predictive of the compared modeling approaches. SVM, RF, and BT modeling techniques were able to successfully predict cyanobacterial abundance, microcystin, and geosmin concentrations <60,000 cells/mL, 2.5 µg/L, and 20 ng/L, respectively. Only Cubist modeling predicted maxima concentrations of cyanobacteria and geosmin; no modeling technique was able to predict maxima microcystin concentrations. Because maxima concentrations are a primary concern for lake and reservoir managers, Cubist modeling may help predict the largest and most noxious concentrations of cyanobacteria and their secondary metabolites.

Predicting behavior of FRP-confined concrete using neuro fuzzy, neural network, multivariate adaptive regression splines and M5 model tree techniques

This paper studies the ability of artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS), multivariate adaptive regression splines (MARS) and M5 Model Tree (M5Tree) techniques to predict ultimate conditions of fiber-reinforced polymer (FRP)-confined concrete. A large experimental test database that consists of over 1000 axial compression tests results of FRP-confined concrete specimens assembled from the published literature was used to train, test, and validate the models. The modeling results show that the ANN, ANFIS, MARS and M5Tree models fit well with the experimental test data. The M5Tree model performs better than the remaining models in predicting the hoop strain reduction factor and strength enhancement ratio, whereas the ANN model provided the most accurate estimates of the strain enhancement ratio. Performances of the proposed models are also compared with those of the existing conventional and evolutionary algorithm models, which indicate that the proposed ANN, ANFIS, MARS and M5Tree models exhibit improved accuracy over the existing models. The predictions of each proposed model are subsequently used to establish the interdependence of critical parameters and their influence on the behavior of FRP-confined concrete, which are discussed in the paper.

Tree Modeling with Real Tree-Parts Examples.

We introduce a 3D tree modeling technique that utilizes examples of real trees to enhance tree creation with realistic structures and fine-level details. In contrast to previous works that use smooth generalized cylinders to represent tree branches, our method generates realistic looking tree models with complex branching geometry by employing an exemplar database consisting of real-life trees reconstructed from scanned data. These trees are sliced into representative parts (denoted as tree-cuts), representing trunk logs and branching structures. In the modeling process, tree-cuts are positioned in space in an intuitive manner, serving as efficient proxies that guide the creation of the complete tree. Allometry rules are taken into account to ensure reasonable relations between adjacent branches. Realism is further enhanced by automatically transferring geometric textures from our database onto tree branches as well as by guided growing of foliage. Our results demonstrate the complexity and variety of trees that can be generated with our method within few minutes. We carry a user study to test the effectiveness of our modeling technique.

A model tree approach to forecasting solar irradiance variability

As the penetration of solar power increases, the variable generation from this renewable resource will necessitate solar irradiance forecasts for utility companies to balance the energy grid. In this study, the temporal irradiance variability is calculated by the temporal standard deviation of the Global Horizontal Irradiance (GHI) at eight sites in the Sacramento Valley and the spatial irradiance variability is quantified by the standard deviation across those same sites. Our proposed artificial intelligence forecasting technique is a model tree with a nearest neighbor option to predict the irradiance variability directly. The model tree technique reduces the mean absolute error of the variability prediction between 10% and 55% compared to using climatological average values of the temporal and spatial GHI standard deviation. These forecasts are made at 15-min intervals out to 180-min. A data denial experiment showed that the addition of surface weather observations improved the forecasting skill of the model tree by approximately 10%. These results indicate that the model tree technique can be implemented in real-time to produce solar variability forecasts to aid utility companies in energy grid management.

Reliability Research of Fiber-Access Devices Monitoring System on Fault Tree

The method of fault tree analysis is an important method for complex system reliability analysis of large scale. Fiber-access devices monitoring system is an important system of electric power communication transmission. Firstly it introduces the method of fault tree analysis, and then use the fault tree modeling technique for the fault tree to model refers to the fault of optical access equipment monitoring system, and carries on the qualitative analysis and quantitative analysis. It provides theoretical basis for system improvement and maintenancs.

Microarchitectural design space exploration made fast

Predictive modeling is an emerging methodology for microarchitectural design space exploration. However, this method suffers from high costs to construct predictive models, especially when unseen programs are employed in performance evaluation. In this paper, we propose a fast predictive model-based approach for microarchitectural design space exploration. The key of our approach is utilizing inherent program characteristics as prior knowledge (in addition to microarchitectural configurations) to build a universal predictive model. Thus, no additional simulation is required for evaluating new programs on new configurations. Besides, due to employed model tree technique, we can provide insights of the design space for early design decisions. Experimental results demonstrate that our approach is comparable to previous approaches regarding their prediction accuracies of performance/energy. Meanwhile, the training time of our approach achieves 7.6–11.8× speedup over previous approaches for each workload. Moreover, the training costs of our approach can be further reduced via instrumentation technique.