Linear Neural Network Models Research Articles

Comparison Between Multiple Linear Regression (MLR) Model and Artificial Neural Network(ANN) Model for the Lean Practices of Manufacturing Industries (SMEs) of Gujarat

Lean Manufacturing practices are basically waste reduction perspective, in which value adding activities are focused over non value adding activities, however there are other advantages like process improvement, muda elimination, muri elimination but its main principal is work improvement through process flow and pull kanban.which finally results in increased productivity and key performance factor improvement like quality, cost, financial impact. Performance measures can only be controlled, if the critical factors are identified and explored among various process parameters. The objective of the paper is to compare the ANN base SPSS model with conceptual model and its linear regression model for lean manufacturing for SMEs. The comparison will gives the closeness of the data obtained for Model development and theory development .The residual indices were measured and equated to predict the performance measures. The operational level factors were explored for different performance measures of lean practices and their synaptic weightage for ANN was determined by activating hidden layer method in SPSS 20.

Kinetic Data Analysis by MLR and ANN Models for Phenol Attenuation in Peat Soil

AbstractThe efficacy of phenol transport from the aqueous phase to peat soil for assessment of the attenuation capacity for migratory phenol in subsurface water pollution was investigated by the application of multiple linear regression (MLR) and artificial neural network (ANN) models. The batch kinetics study was performed, which revealed that the Freundlich isotherm model fits reasonably well with experimental results. A maximum value of 43% phenol removal efficiency was achieved in 200 g/L of soil, an initial concentration of phenol of 10 mg/L, and an equilibration time of 6 h. A sum total of 270 laboratory batch adsorption tests were conducted, and the results were applied in MLR and ANN models. Some of the influencing factors, such as pH, initial concentration, mass of soil, contact time, and so forth, on removal of sorbate by peat were also investigated in the present research. The experimental results exhibit a reasonable goodness of fit [higher coefficient of determination, R2, and lower root-mean...

On the Ergodicity of General Mixture of Linear Autoregressive Time Series

In this paper, a general framework of mixture time series is introduced as General Mixture of Linear Autoregressive. This framework contains well known mixture time series such as Mixture Autoregressive, Local Linear Global Neural Network and Logistic Mixture Autoregressive models. The sufficient conditions for geometric ergodicity of such model are derived.

College Towns, Vacation Spots, and Tech Hubs: Using Geo-Social Media to Model and Compare Locations

In this paper, we explore the potential of geo-social media to construct location-based interest profiles to uncover the hidden relationships among disparate locations. Through an investigation of millions of geo-tagged Tweets, we construct a per-city interest model based on fourteen high-level categories (e.g., technology, art, sports). These interest models support the discovery of related locations that are connected based on these categorical perspectives (e.g., college towns or vacation spots) but perhaps not on the individual tweet level. We then connect these city-based interest models to underlying demographic data. By building multivariate multiple linear regression (MMLR) and neural network (NN) models we show how a location's interest profile may be estimated based purely on its demographics features.

Neural network forecast of daily pollution concentration using optimal meteorological data at synoptic and local scales

We present a simple neural network and data pre–selection framework, discriminating the most essential input data for accurately forecasting the concentrations of PM10, based on observations for the years between 2002 and 2006 in the metropolitan region of Lisbon, Portugal. Starting from a broad panoply of different data sets collected at several air quality and meteorological stations, a forward stepwise regression procedure is applied enabling to automatically identify the most important variables for predicting the pollutant and also to rank them in order of importance. The importance of this variable ranking is discussed, showing that it is very sensitive to the urban location where measurements are obtained. Additionally, the importance of Circulation Weather Types is highlighted, characterizing synoptic scale circulation patterns and the concentration of pollutants. We then quantify the performance of linear and non–linear neural network models when applied to PM10 concentrations. In the light of contradictory results of previous studies, our results show no clear superiority for the case studied of non–linear models over linear models. While all models show similar predictive performances, we find important differences in false alarm rates and demonstrate the importance of removing weekly cycles from input variables.

A four-stage hybrid model for hydrological time series forecasting.

Hydrological time series forecasting remains a difficult task due to its complicated nonlinear, non-stationary and multi-scale characteristics. To solve this difficulty and improve the prediction accuracy, a novel four-stage hybrid model is proposed for hydrological time series forecasting based on the principle of ‘denoising, decomposition and ensemble’. The proposed model has four stages, i.e., denoising, decomposition, components prediction and ensemble. In the denoising stage, the empirical mode decomposition (EMD) method is utilized to reduce the noises in the hydrological time series. Then, an improved method of EMD, the ensemble empirical mode decomposition (EEMD), is applied to decompose the denoised series into a number of intrinsic mode function (IMF) components and one residual component. Next, the radial basis function neural network (RBFNN) is adopted to predict the trend of all of the components obtained in the decomposition stage. In the final ensemble prediction stage, the forecasting results of all of the IMF and residual components obtained in the third stage are combined to generate the final prediction results, using a linear neural network (LNN) model. For illustration and verification, six hydrological cases with different characteristics are used to test the effectiveness of the proposed model. The proposed hybrid model performs better than conventional single models, the hybrid models without denoising or decomposition and the hybrid models based on other methods, such as the wavelet analysis (WA)-based hybrid models. In addition, the denoising and decomposition strategies decrease the complexity of the series and reduce the difficulties of the forecasting. With its effective denoising and accurate decomposition ability, high prediction precision and wide applicability, the new model is very promising for complex time series forecasting. This new forecast model is an extension of nonlinear prediction models.

An Effective Method to Evaluate the Scientific Research Projects

Abstract The evaluation of the scientific research projects is an important procedure before the scientific research projects are approved. The BP neural network and linear neural network are adopted to evaluate the scientific research projects in this paper. The evaluation index system with 12 indexes is set up. The basic principle of the neural network is analyzed and then the BP neural network and linear neural network models are constructed and the output error function of the neural networks is introduced. The Matlab software is applied to set the parameters and calculate the neural networks. By computing a real-world example, the evaluation results of the scientific research projects are obtained and the results of the BP neural network, linear neural network and linear regression forecasting are compared. The analysis shows that the BP neural network has higher efficiency than the linear neural network and linear regression forecasting in the evaluation of the scientific research projects problem. The method proposed in this paper is an effective method to evaluate the scientific research projects.

Prediction Models for Railway Freight Volume Based on Artificial Neural Networks

The current China railway freight transport has always been faced with the situation of limited transport resources. Many relative studies have been done to solve the problem of resource shortage. And railway freight volume prediction is the basis of all these studies. With accurate volume prediction, railway freight transport administrations can precisely allocate the transport resources, such as wagons and locomotives. In order to overcome the limitations of traditional prediction methods, in this study, we design four artificial neural network models for prediction, including BP neural network model, linear neural network model, RBF neural network model and generalized regression neural network model. The results of simulation and comparison show that all these models can reach high prediction accuracy and generalized regression neural network has both higher prediction accuracy and better curve fitting capacity compared with other models.

Atmospheric Pollution: Statistical Models of Predicting SO<sub>2</sub> Concentration in Shanghai

Air quality has been deteriorated seriously in Shanghai as a result of urbanization and modernization. Meteorological conditions affect air pollution levels in the urban atmosphere significantly due to their important role in transport and dilution of the pollutants. This paper aims to investigate usability of statistical methods for prediction of SO2concentration. Multiple linear regression (MLR) model and artificial neural network (ANN) model were used to predict next days SO2concentrations. The calculatedR2values were 0.56 and 0.55 for MLR and ANN model, respectively. This result shows the MLR analysis, providing simple and faster application facilities, may also be useful for modeling the impacts of meteorological factors on SO2levels besides the time consuming ANN model.

Dynamic Reaction Fitting for Gaseous Formaldehyde Detector with Unmeasured Data Prediction

This paper focuses on the sparsity and artifacts of the dynamic reaction of gaseous formaldehyde detector, which results in a certain distortion in the fitting process. A new method based on linear regressive model with unmeasured data prediction using linear neural network is proposed . Linear neural network method has simple inferring rules and strong predictive capability for linear relationships. This predictive data can be obtained using the sequence near the end as the input signal of trained linear neural network model. Based on the extended data, a linear regressive model is introduced to fit the response curve of gaseous formaldehyde detector. Experimental results have shown that the presented method is valid and effective .

ANN multiplexing model of drugs effect on macrophages; theoretical and flow cytometry study on the cytotoxicity of the anti-microbial drug G1 in spleen

Multiplexed biological assays provide multiple measurements of cellular parameters in the same test. In this work, we have trained and tested an Artificial Neural Network (ANN) model for the first time, in order to perform a multiplexing prediction of drugs effect on macrophage populations. In so doing, we have used the TOPS-MODE approach to calculate drug molecular descriptors and the software STATISTICA to seek different ANN models such as: Linear Neural Network (LNN), Radial Basis Function (RBF), Probabilistic Neural Networks (PNN) and Multi-Layer Perceptrons (MLP). The best model found was the LNN, which correctly classified 8258 out of 9000 (Accuracy=93.0%) multiplexing assay endpoints of 7903 drugs (including both training and test series). Each endpoint corresponds to one out of 1418 assays, 36 molecular or cellular targets, 46 standard type measures, in two possible organisms (human and mouse). Secondly, we have determined experimentally, for the first time, the values of EC50=11.41μg/mL and Cytotoxicity=27.1% for the drug G1 over Balb/C mouse spleen macrophages using flow cytometry. In addition, we have used the LNN model to predict the G1 activity in 1265 multiplexing assays not measured experimentally (including 152 cytotoxicity assay endpoints). Both experimental and theoretical results point out a low macrophage cytotoxicity of G1. This work breaks new ground for the ‘in silico’ multiplexing screening of large libraries of compounds. The results obtained are very significant because they complement the immunotoxicology studies of this important anti-microbial/anti-parasite drug.

Learning with Slight Forgetting Optimizes Sensorimotor Transformation in Redundant Motor Systems

Recent theoretical studies have proposed that the redundant motor system in humans achieves well-organized stereotypical movements by minimizing motor effort cost and motor error. However, it is unclear how this optimization process is implemented in the brain, presumably because conventional schemes have assumed a priori that the brain somehow constructs the optimal motor command, and largely ignored the underlying trial-by-trial learning process. In contrast, recent studies focusing on the trial-by-trial modification of motor commands based on error information suggested that forgetting (i.e., memory decay), which is usually considered as an inconvenient factor in motor learning, plays an important role in minimizing the motor effort cost. Here, we examine whether trial-by-trial error-feedback learning with slight forgetting could minimize the motor effort and error in a highly redundant neural network for sensorimotor transformation and whether it could predict the stereotypical activation patterns observed in primary motor cortex (M1) neurons. First, using a simple linear neural network model, we theoretically demonstrated that: 1) this algorithm consistently leads the neural network to converge at a unique optimal state; 2) the biomechanical properties of the musculoskeletal system necessarily determine the distribution of the preferred directions (PD; the direction in which the neuron is maximally active) of M1 neurons; and 3) the bias of the PDs is steadily formed during the minimization of the motor effort. Furthermore, using a non-linear network model with realistic musculoskeletal data, we demonstrated numerically that this algorithm could consistently reproduce the PD distribution observed in various motor tasks, including two-dimensional isometric torque production, two-dimensional reaching, and even three-dimensional reaching tasks. These results may suggest that slight forgetting in the sensorimotor transformation network is responsible for solving the redundancy problem in motor control.

Support Vector Regression Model of Chlorophyll-<i>a</i> during Spring Algal Bloom in Xiangxi Bay of Three Gorges Reservoir, China

To study the relationship between chlorophyll-a and environmental variables during spring algal bloom in Xiangxi Bay of Three Gorges Reservoir, the support vector regression (SVR) model was established. In surveys, 11 stations have been investigated and 264 samples were collected weekly from March 4 to May 13 in 2007 and February 16 to May 10 in 2008. The parameters in SVR model were optimized by leave one out cross validation. The squared correlation coefficient R2 and the cross validated squared correlation coefficient Q2 of the optimal SVR model are 0.8202 and 0.7301, respectively. Compared with stepwise multiple linear regression and back propagation artificial neural network models using external validation, the SVR model has been shown to perform well for regression with the predictive squared correlation coefficient R2pred value of 0.7842 for the test set.

Comparison of the 3-phase segmented linear regression and artificial neural network models to predict broiler hatchability

SUMMARY The objective of this study was to compare the performance of an artificial neural network (ANN) model and a 3-phase segmented linear regression model to describe the relationship between flock age and hatchability in broiler breeder flocks. The predictive quality of these models was tested for an external validation set of 14 wk, randomly chosen from 39 wk. The accuracy of the models was determined by the r2 value, mean square error, bias, and Theil’s U-statistic parameters. The r2 values of the 3-phase segmented linear regression and ANN models were 0.4003 and 0.9984, respectively. Therefore, the ANN produced more accurate predictions of hatchability than the 3-phase segmented linear regression model. We conclude, based on the results of this study in commercial broiler breeder flocks, that hatchability is a function of flock age and that the relationship can be described by an ANN model.

A Comparison between the Linear Neural Network Method and the Multiple Linear Regression Method in the Modeling of Continuous Data

Both linear neural network and multiple linear regression models can be used for multi-factor analysis and forecasting, but the data of the multiple linear regression model are required to meet such conditions as independence and normality, while the data of the linear neural network are only required to have a linear relationship. This article uses the same set of data to establish respectively a linear neural network model and a multiple linear regression model, compares the abilities of fitting and forecasting of the two kinds of models, and consequently, comes to the conclusion that the linear neural network method has a stronger fitting ability and a more stable ability of prediction so that it can be further applied and promoted in the analyzing and forecasting of continuous data factors.

QSAR study on cytotoxic activity (against KB cells) of some hederagenin diglycosides using support vector regression

Quantitative structure-activity relationship (QSAR) study on the cytotoxic activity (against KB cells) of 19 hederagenin diglycosides was performed by using spatial and electronic descriptors based on support vector regression (SVR) techniques. The predictive power of the models was verified with the leave one out cross validation (LOOCV) test and independent test methods. For the LOOCV test, the cross validation squared correlation coefficient Q2 value for optimal SVR model was 0.8827. Compared with stepwise multiple linear regression (MLR) and back propagation artificial neural network (BPANN) models, the SVR model was the most powerful with a square of predictive correlation coefficient \({R_{pred}^2}\) of 0.7285 for the test set, which indicates that the SVR model has better substantially predictive ability and be a useful and powerful tool to construct the QSAR model.

Graybox and adaptative dynamic neural network identification models to infer the steady state efficiency of solar thermal collectors starting from the transient condition

This paper deals with the development of methods for non steady state test of solar thermal collectors. Our goal is to infer performances in steady-state conditions in terms of the efficiency curve when measures in transient conditions are the only ones available. We take into consideration the method of identification of a system in dynamic conditions by applying a Graybox Identification Model and a Dynamic Adaptative Linear Neural Network (ALNN) model. The study targets the solar collector with evacuated pipes, such as Dewar pipes. The mathematical description that supervises the functioning of the solar collector in transient conditions is developed using the equation of the energy balance, with the aim of determining the order and architecture of the two models. The input and output vectors of the two models are constructed, considering the measures of 4 days of solar radiation, flow mass, environment and heat-transfer fluid temperature in the inlet and outlet from the thermal solar collector. The efficiency curves derived from the two models are detected in correspondence to the test and validation points. The two synthetic simulated efficiency curves are compared with the actual efficiency curve certified by the Swiss Institute Solartechnik Pufung Forschung which tested the solar collector performance in steady-state conditions according to the UNI-EN 12975 standard. An acquisition set of measurements of only 4 days in the transient condition was enough to trace through a Graybox State Space Model the efficiency curve of the tested solar thermal collector, with a relative error of synthetic values with respect to efficiency certified by SPF, lower than 0.5%, while with the ALNN model the error is lower than 2.2% with respect to certified one.

SIMULATION OF MONTHLY STREAMFLOWS USING LINEAR REGRESSION AND ANN MODELS

ABSTRACT Streamflow simulation is essential for planning, designing and operation of water resources projects. In the present study, monthly streamflows during monsoon period are simulated using linear regression and artificial neural network models. The gauging sites of Mancherial, Perur and Polavaram of Godavari basin of India are selected for the present study. The study reveals that the performance of linear regression models in the simulation of streamflow is satisfactory and improvement in the performance has been observed using optimal neural network architectures. The linear regression model may be adopted for the simulation of streamflows at Polavaram gauging site where the flows do not exhibit much of non-linearity and ANN models at Peruru and Mancherial gauging sites using the streamflows of upstream gauging sites of the basin.

Distributed Representations Accelerate Evolution of Adaptive Behaviours

Animals with rudimentary innate abilities require substantial learning to transform those abilities into useful skills, where a skill can be considered as a set of sensory–motor associations. Using linear neural network models, it is proved that if skills are stored as distributed representations, then within-lifetime learning of part of a skill can induce automatic learning of the remaining parts of that skill. More importantly, it is shown that this “free-lunch” learning (FLL) is responsible for accelerated evolution of skills, when compared with networks which either 1) cannot benefit from FLL or 2) cannot learn. Specifically, it is shown that FLL accelerates the appearance of adaptive behaviour, both in its innate form and as FLL-induced behaviour, and that FLL can accelerate the rate at which learned behaviours become innate.

Neural network modeling of ecosystems: A case study on cabbage growth system

A deep understanding on the intrinsic mechanism is required to develop a highly specialized mechanistic model for ecosystem dynamics. However, it is usually hard to do for most of the ecological and environmental problems, because of the lack of a consistent theoretical background. Neural networks are universal and flexible models for linear and non-linear systems. This paper aimed to modeling an ecosystem using neural network models and the conventional model, and assessing their effectiveness in the dynamic simulation of ecosystem. Elman neural network model, linear neural network model, and linear ordinary differential equation were developed to simulate the dynamics of Chinese cabbage growth system recorded in the field. Matlab codes for these neural network models were given. Sensitivity analysis was conducted to detect the robustness of these models. The results showed that Elman neural network model could simulate the multivariate non-linear system at the desired accuracy. Linear neural network model may simulate such a non-linear system only in certain conditions. Conventional linear ordinary differential equation yielded divergent outputs in the dynamic simulation of the multivariate non-linear system. Sensitivity analysis indicated that the learning rate influenced the simulation performance of linear neural network model. Transfer function of the second layer in Elman neural network model would influence the simulation performance of this model, but little influence was produced when other functions were changed. Elman neural network was proven to be a robust model. Sensitivity analysis showed that the different choices of functions and parameters in neural network model would influence the performance of simulation. Sensitivity analysis is therefore powerful to detect the robustness and stability of neural network models.