Hypsometric relationship of Schizolobium parahyba var. amazonicum in plantations integrated with livestock in eastern Amazonia: applications of different modeling methods

Regression and ANN models for predicting MOR and MOE of heat-treated fir wood

The study focuses on computing the optimized foot profile for a walking leg mechanism using artificial neural network (ANN), genetic algorithm, and regression approaches. The technique adopted in this work is the benchmark approach and acts as a tool for complex problems. A mathematical model using regression and ANN is developed for the 8-link coplanar mechanism. Optimum link lengths are obtained to minimize the objective function (error). The output response is the foot length with a minimum foot height of 124 mm for obstacle clearance. A neural network is designed with seven neurons (one neuron/link) in the input layer. Optimum neurons in the hidden layer are determined based on the output obtained through simulation. A single neuron is used to represent the foot profile length at the output layer. The foot lengths obtained from the regression model and ANN are compared and validated with a genetic algorithm for the data sets of 100, 200, 300, 400, and 500. Simulation studies of the walking leg mechanism revealed a difference of 19%, 22.4%, and 5.23% in the foot profile by ANN and mathematical, ANN and regression model, and mathematical and regression approach respectively. This paper reveals that different approaches viz., ANN, mathematical and regression models generate dissimilar foot profiles.

The term soft computing refers to a family of techniques consisting of methods and procedures based on fuzzy logic, evolutionary computing, artificial neural networks, probabilistic reasoning, rough sets, chaotic computing. With the discovery that the Web is structured according to social networks exhibiting the small world property, the idea of using taxonomy principles has appeared as a complementary alternative to traditional keyword searching. One technique which has emerged from this principle was the “web-as-brain” metaphor. It is yielding new, associative, artificial neural networks- (ANN-) based retrieval techniques. The present paper proposes a unified formal framework for three major methods used for Web retrieval tasks: PageRank, HITS, I2R. The paper shows that these three techniques, albeit they stem originally from different paradigms, can be integrated into one unified formal view. The conceptual and notational framework used is given by ANNs and the generic network equation. It is shown that the PageRank, HITS and I2R methods can be formally obtained from the generic equation as different particular cases by making certain assumptions reflecting the corresponding underlying paradigm. The unified formal view sheds a new light upon the understanding of these methods: it may be said that they are only seemingly different from each other, they are particular ANNs stemming from the same equation and differing from one another in whether they are dynamic (a page’s importance varies in time) or static (a page’s importance is constant in time), and in the way they connect the pages to each other. The paper also gives a detailed mathematical analysis of the computational complexity of WTA-based IR techniques using the I2R method for illustration. The importance of this analysis consists in that it shows that (i) intuition may be misleading (contrary to intuition, a WTA-based algorithm yielding circles is not always “hard”), and (ii) this analysis can serve as a model that may be followed in the analysis of other methods.

Artificial neural network modelling of the amount of separately-collected household packaging waste

Purpose This paper aims to propose the artificial neural network (ANN) and regression models for the estimation of the thread consumption at multilayered seam assembly stitched with lock stitch 301. Design/methodology/approach In the present study, the generalized regression and neural network models are developed by considering the fabric types: woven, nonwoven and multilayer combination thereof, with basic sewing parameters: sewing thread linear density, stitch density, needle count and fabric assembly thickness. The network with feed-forward backpropagation is considered to build the ANN, and the training function trainlm of MATLAB software is used to adjust weight and basic values according to the optimization of Levenberg Marquardt. The performance of networks measured in terms of the mean squared error and the layer output is set according to the sigmoid transfer function. Findings The proposed ANN and regression model are able to predict the thread consumption with more accuracy for multilayered seam assembly. The predictability of thread consumption from available geometrical models, regression models and industrial empirical techniques are compared with proposed linear regression, quadratic regression and neural network models. The proposed quadratic regression model showed a good correlation with practical thread consumption value and more accuracy in prediction with an overall 4.3% error, as compared to other techniques for given multilayer substrates. Further, the developed ANN network showed good accuracy in the prediction of thread consumption. Originality/value The estimation of thread consumed while stitching is the prerequisite of the garment industry for inventory management especially with the introduction of the costly high-performance sewing thread. In practice, different types of fabrics are stitched at multilayer combinations at different locations of the stitched product. The ANN and regression models are developed for multilayered seam assembly of woven and nonwoven fabric blend composition for better prediction of thread consumption.

A Comparison of the Performances of an Artificial Neural Network and a Regression Model for GFR Estimation

The purpose of this paper is to forecast housing prices in Ankara, Turkey using the artifi cial neural networks (ANN) approach. The data set was collected from one of the biggest real estate web pages during April 2013. A three-layer (input layer – one hidden layer – output layer) neural network is designed with 15 different inputs to forecast the future housing prices. The proposed model has a success rate of 78%. The results of this paper would help property investors and real estate agents in developing more effective property pricing management in Ankara. We believe that the artifi cial neural networks (ANN) proposed here will serve as a reference for countries that develop artifi cial neural networks (ANN) method-based housing price determination in future. Applying the artifi cial neural networks (ANN) approach for estimation of housing prices is relatively new in the fi eld of housing economics. Moreover, this is the fi rst study that uses the artifi cial neural networks (ANN) approach for analyzing the housing market in Ankara/Turkey.

Increasing temperature from climate change can bring a number of different risks such as more droughts and heat waves, and increasing sea level rise. Assessment of climate change with future scenarios is essential to adapt these impacts. To provide climate change information through the outputs of general circulation models at finer resolution, a reliable and accurate downscaling model has always been of great interest. Meeting this need, artificial neural network (ANN) has been commonly employed in downscaling for nonlinear models. Extreme learning machine (ELM), a recently developed ANN, is an efficient learning algorithm for generalized single hidden layer feedforward neural networks. In light of its simple learning algorithm, we introduced a useful approach to combine the stepwise feature selection method into ELM for temperature downscaling, as stepwise ELM (SWELM), since model complexity and computational time consumption of a traditional ANN impedes application of stepwise feature selection. This SWELM is able to identify the most influential predictors in a dataset and use them to train a nonlinear model while removing the irrelevant ones. The ELM and SWELM as well as regular ANN were tested in a simulation study. Results indicated that ELM even with randomness of weights and biases in the nodes of input and hidden layers better performed than did ANN. Also, SWELM presents a capability to select the influential predictors and remove the unrelated variables. A case study with downscaling temperature of Wisconsin, USA, showed that ELM was a comparable alternative to ANN. SWELM outperformed the ANN algorithm for temperature downscaling and sometimes predicted the temperature increase larger than did others for future scenarios. The current study of temperature downscaling with the statistical tool allows assessing the possible impacts of climate change in a local scale and some developing countries where sophisticate research cannot be eligible.

Prediction of event-based stormwater runoff quantity and quality by ANNs developed using PMI-based input selection

Wind gusts, and in particular intense gusts, are societally relevant but extremely challenging to forecast. This study systematically assesses the skill enhancement that can be achieved using artificial neural networks (ANNs) for forecasting of wind gust occurrence and magnitude. Geophysical predictors from the ERA5 reanalysis are used in conjunction with an autoregressive term in regression and ANN models with different predictors, and varying model complexity. Models are derived and assessed for the warm (April–September) and cold (October–March) seasons for three high passenger volume airports in the United States. Model uncertainty is assessed by deriving models for 1000 different randomly selected training (70%) and testing (30%) subsets. Gust prediction fidelity in independent test samples is critically dependent on inclusion of an autoregressive term. Gust occurrence probabilities derived using five-layer ANNs exhibit consistently higher fidelity than those from regression models and shallower ANNs. Inclusion of the autoregressive term and increasing the number of hidden layers in ANNs from 1 to 5 also improve the model performance for gust magnitudes (lower RMSE, increased correlation, and model standard deviations that more closely approximate observed values). Deeper ANNs (e.g., 20 hidden layers) exhibit higher skill in forecasting strong (17–25.7 m s−1) and damaging (≥25.7 m s−1) wind gusts. However, such deep networks exhibit evidence of overfitting and still substantially underestimate (by 50%) the frequency of strong and damaging wind gusts at the three airports considered herein. Significance Statement Improved short-term forecasting of wind gusts will enhance aviation safety and logistics and may offer other societal benefits. Here we present a rigorous investigation of the relative skill of models of wind gust occurrence and magnitude that employ different statistical methods. It is shown that artificial neural networks (ANNs) offer considerable skill enhancement over regression methods, particularly for strong and damaging wind gusts. For wind gust magnitudes in particular, application of deeper learning networks (e.g., five or more hidden layers) offers tangible improvements in forecast accuracy. However, deeper networks are vulnerable to overfitting and exhibit substantial variability with the specific training and testing data subset used. Also, even deep ANNs reproduce only half of strong and damaging wind gusts. These results indicate the need for future work to elucidate the dynamical mechanisms of intense wind gusts and advance solutions to their prediction.

This paper describes the development of regression and artificial neural network (ANN) models to determine the 28-day compressive and tensile strength of engineered cementitious composite (ECC) based on the mix design parameters. One hundred eighty ECC mixtures having variable mix designs were obtained from pervious experiments. Factors influencing the strengths were examined to determine the appropriate parameters for the ANN models. The optimized input parameters using training and development of ANN models were used to formulate the regression models. The ANN and regression models were tested with new sets of data for performance validation. Based on the good agreement and other statistical performance parameters, optimized ANN and regression models capable of predicting the strengths of ECC mixtures (using arbitrary mix design parameters) were developed and suggested for practical applications. ANN and regression models demonstrated excellent predictive ability showing predicted experimental strength ratio ranging between 0.95 and 1.00.

The work presented in this paper is an investigation of the prediction of amplitudes of the specific grinding force components. An improved method for artificial neural networks (ANNs) establishment is proposed here allowing accurate prediction of specific normal and tangential grinding forces. This method can determine the optimal set of inputs to be used for these ANN. This set of inputs is composed of significant factors and interactions among factors that could possibility offer the best learning and generalization of ANNs simultaneously. A 48-run experimental design (MED) is used in this research to train the ANNs and a total of 81 experiments are conducted to test the generalization performances of ANNs. Results have indicated that the developed ANNs show low deviations from the training data, and acceptable deviations from the testing data. In addition, the accuracies of these ANNs are found to be significantly better than those of other approaches used for modelling of the specific grinding force components. These approaches use regression models, power models, genetic algorithms or the common ANNs for which only factors of the MED are usually used in the input layer.

Real-time nondestructive monitoring of Common Carp Fish freshness using robust vision-based intelligent modeling approaches

Artificial Neural Networks (ANN) have been proven applicable for updating finite-element (FE) baseline model and structural damage assessment. Most ANN-based damage identification methods use natural frequencies and mode shapes as input layer, limiting their application to quantifying single symmetrical damage in small structures. However, getting higher modal information of a structure is a crucial challenge in practice. As of late, researchers began utilizing mode shape derivatives as input layer in ANN to defeat the challenges for damage assessment in real-life structures. This study, therefore, proposes an ANN-based damage assessment method that employs the change in the first mode shape slope (CFMSS) damage index (DI) as input layer in ANN. For single-damage scenarios, the CFMSS-based DI has been able to detect, locate, and quantify the damage. For multiple-damage scenarios, the DI and corresponding stiffness reduction (SR) are fit as input and output layers, respectively, in ANN to measure the damage severity. Structural damage intensity is indicated as rate of decrease in story stiffness compared to baseline model. The efficiency of the proposed damage identification method is demonstrated through a nine-story numerical shear frame model and an experimental test on a three-story steel shear frame model.

Recently statistical model especially forecasting model has developed to soft model. The model is more computerize in line to computer development and it is not based on strict rules, such as it has to fulfil classical or soft assumption. The model is called as soft statistical model. By using soft model and soft assumption, there are many models can be constructed, such as Artificial Neural Network (ANN) or Multi Layers Perceptions (MLP). There are three layers in ANN, it is called input, hidden and output layer. The optimum weight of each layer is processed using back propagation approach. In this research, ANN model – especially Neural Fuzzy Regression (NFR) model – is applied to find best forecasting model, specifically forecasting model of the stock price. The data is the stock price of a mining sector emitted and the exchange rate US$ to IDR from January 2015 until February 2019. The Data is collected from publication of Indonesia Stock Exchange and Indonesia Central Bank. The stock price shows positive trends recently and there is a correlation between the stock price and the exchange rate. Based on autocorrelation function, there are four previous data that have significant relationship with the current data. NFR model has five nodes in input layer (four lag time and exchange rate), some nodes in hidden layer and a node in output layer. The best model is model with five nodes as input, seven nodes in hidden layer and an output. The model has accuracy of MSE 34.0850, MAPE 2.9026, and MAD 28.7377.