Optimal Regression Coefficients Research Articles

Artificial intelligence-based approaches to evaluate and optimize phytoremediation potential of in vitro regenerated aquatic macrophyte Ceratophyllum demersum L.

Water bodies or aquatic ecosystem are susceptible to heavy metal accumulation and can adversely affect the environment and human health especially in underdeveloped nations. Phytoremediation techniques of water bodies using aquatic plants or macrophytes are well established and are recognized as eco-friendly world over. Phytoremediation of heavy metals and other pollutants in aquatic environments can be achieved by using Ceratophyllum demersum L. - a well-known floating macrophyte. In vitro regenerated plants of C. demersum (7.5g/L) were exposed to 24, 72, and 120h to 0, 0.5, 1.0, 2.0, and 4.0mg/L of cadmium (CdSO4·8H2O) in water. Results revealed significantly different relationship in terms of Cd in water, Cd uptake by plants, bioconcentration factor (BCF), and Cd removal (%) from water. The study showed that Cd uptake by plants and BCF values increased significantly with exposure time. The highest BCF value (3776.50) was recorded for plant samples exposed to 2mg/L Cd for 72h. Application of all Cd concentrations and various exposure duration yielded Cd removal (%) between the ranges of 93.8 and 98.7%. These results were predicted through artificial intelligence-based models, namely, random forest (RF), extreme gradient boosting (XGBoost), and multilayer perceptron (MLP). The tested models predicted the results accurately, and the attained results were further validated via three different performance metrics. The optimal regression coefficient (R2) for the models was recorded as 0.7970 (Cd water, mg/L), 0.9661 (Cd plants, mg/kg), 0.9797 bioconcentration factor (BCF), and 0.9996 (Cd removal, %), respectively. These achieved results suggest that in vitro regenerated C. demersum can be efficaciously used for phytoremediation of Cd-contaminated aquatic environments. Likewise, the proposed modeling of phytoremediation studies can further be employed more comprehensively in future studies aimed at data prediction and optimization.

Device Free Localization Using Prior Correction-Based Logistic Regression

In this letter, we focus on a device free localization (DFL) system which is based on logistic regression model using received signal strength (RSS) measurements in each link as training and testing data. To reduce the computational complexity of the logistic regression model prediction and improve the localization accuracy by reduction of the redundant training features, we obtain principal component using the measured RSS data through principal component analysis (PCA). Moreover, to overcome the imbalanced data problem, the number of training data belonging to major class is decreased through the sampling technique based on the distance between the reference points. The modified logistic regression model is derived from Bayesian formula based prior correction method due to the change in the ratio of the major and the minor classes in sampled training data. By substituting the transformed principal components of testing dataset to the probability model with the optimal regression coefficients, the localization probabilities of a target are computed at every reference point.

Inversion analysis of the Three Dimension Geostress Field in A Tunnel

The initial geostress field is an important basis for the design and stability analysis of geotechnical engineering. The geostress field of the whole engineering area must be calculated by inversion analysis with the limited geostress data, because there are few measured geostress points. Based on the geological background of a tunnel in northern Hebei province and considering the influence of regional highly weathered, moderately weathered and slightly weathered stratum on the stress field, the 3D finite difference method and multiple linear regression method were used to calculate the three-dimensional geological model of the study region to get the optimal regression coefficient. It is found by comparing the regression value with the field measured value that the two values are equal in magnitude and close in direction, indicating that the geostress field calculated by regression is reasonable, then the distribution law of the geostress field in the study area is further analyzed on this basis. The results show that there is a stress concentration in the tunnel center, z peaking at 50m below the center, and decreasing in the left and right direction with a large stress gradient then getting out of the area of stress concentration and tending to be stable. The study area is a medium geostress area, based on gravity stress, the control effect of gravity stress field is gradually enhanced with the increase of buried depth. The research results have important value in the design and construction of tunnel in special geological conditions.

Determination of plastic properties of weld metals using an optimal regression of tensile and hardness test data

A new method is proposed to obtain the coefficients of plastic constitutive laws of weld metal via tensile test and hardness test. In this method, a general expression between hardness ratio and flow stress model coefficients is established. The predicted flow stresses for the differential areas are integrated over the entire section of tensile specimen to predict the tensile load. The optimal regression coefficients that minimize the difference between the predicted and experimental loads are derived using the Nelder-Mead simplex direction search algorithm. The constraint for the diffuse necking strain is applied in the convergence procedure to avoid the ‘mystical’ material property shown as non-unique solution. The material constants obtained from the proposed method are independent of the width of tensile specimen. The difference of tensile loads between finite element analysis results using the determined material constants and the experiment was less than 8 %. As the derived material constants were shown to be consistent and reliable, the methodology can be applied to general line welded parts.

Diagnostic and prognostic analysis of oil and gas pipeline with allowable corrosion rate in Niger Delta Area, Nigeria

This paper presents diagnostic and prognostic analysis of oil and gas pipeline industries with allowable corrosion rate using artificial neural networks approach. The results revealed sand deposit, carbon dioxide (CO2) partial pressure, pipe age, diameter and length, temperature, flow velocity of the fluid, fluid pressure, chloride contents and pH value of its environment as the relevant parameters affecting corrosion of oil and gas pipeline in this region. Condition prediction of steel pipes used for the transmission of oil and gas varies 0.02 mm/yr to 0.10 mm/yr. The training of the neural network was performed using Levenberg-Marquardt algorithm and optimal regression coefficient was equal to 0.99, for the network 10-40-1. Also, the results show a remarkable agreement with the field measurement. A corrosion severity level of two (0.01 mm/yr to 0.10 mm/yr) oil and gas pipelines was established from the analysis.Keywords: Artificial neural network, Levenberg-Marquardt algorithm, condition prediction, oil and gas data

Whale–crow optimization (WCO)-based Optimal Regression model for Software Cost Estimation

Software Cost Estimation (SCE) is the emerging concern of the software companies during the development phase of the software, as it requires effort and cost factors for modelling the software. These factors are modelled using the Artificial Intelligence models, which seem to be less accurate and non-reliable by increasing the risk factor of the software projects. Thus, for estimating the software cost, meta-heuristics are employed. This paper proposes an algorithm, termed as whale–crow optimization (WCO) algorithm, which is the integration of the whale optimization algorithm (WOA) and the crow search algorithm (CSA). The main function of the WCO algorithm is to determine the Optimal Regression coefficients for the regression models, such as the Linear Regression model and the Kernel Logistic Regression model, to develop an Optimal Regression model to estimate the software cost. The experimentation is carried out using four datasets taken from the Promise software engineering repository to perform effective performance analysis. Analysis is carried out regarding the mean magnitude of relative error (MMRE) that proves that the proposed method of SCE is effective, attaining the average MMRE at a rate of 0.2442 for the proposed Linear Regression model and 0.2692 for the proposed Kernel Regression model.

The association between paralytic side and health-related quality of life in facial palsy: a cross-sectional study of the Korea National Health and Nutrition Examination Survey (2008\u20132012)

BackgroundFacial palsy is known to have correlations with low level of quality of life. However, little is known about the association between preference based health-related quality of life (HRQoL) and paralytic side of facial palsy.MethodsThis study used Korea National Health and Nutrition Examination Survey (KNHANES, 2008–2012) data, only when the facial palsy examination had been included in the survey contents. Hierarchical regression analyses were used to obtain optimal regression coefficients in the association between paralytic side of the facial palsy and HRQoL measured by EuroQoL-5 Dimension (EQ-5D). We also analyzed the association between the deteriorated domains of EQ-5D and facial palsy in both subgroups by using multiple logistic regression models.ResultsWe included the data of 28,106 participants aged ≥19 years who were examined as facial palsy according to House-Brackmann score and completed EQ-5D questionnaire in KNHANES 2008–2012. The mean EQ-5D score was significantly low and percentages of deteriorated numbers in its domains were significantly high in facial palsy group.ConclusionsThese results show that, after adjusting for confounding variables, left facial palsy is associated with impaired HRQoL compared with right-sided palsy. Among the domains of EQ-5D, only ‘self-care’ domain was directly affected by the disease in left facial palsy patients. These findings could be used in developing model and conducting analyses of economic evaluation about facial palsy interventions.

An adaptive line search scheme for approximated nuclear norm based matrix regression

Abstract Face recognition with partial occlusion is common in the real world and has become a hot topic in the pattern recognition research. Recently, nuclear norm based matrix regression model (NMR) has been shown a great potential in dealing with the structural noise. And yet, it uses Alternating Direction Method of Multipliers (ADMM) to achieve the optimal regression coefficients, which needs to bring into an auxiliary variable and only exploits the convexity of NMR. Comparing with ADMM, the gradient based methods are simpler. To take advantage of these methods, this paper considers the Approximated NMR model with Elastic-net regularization (ANMRE). By virtue of the soft-thresholding (shrinkage) operator, singular value shrinkage operator and strong convexity of ANMRE, the dual problem of ANMRE (DANMRE) is derived and a crucial result is obtained: the primal optimal solution of ANMRE can be converted as the matrix function associated with the dual optimal solution. Leveraging the differentiability of DANMRE, an adaptive line search scheme is developed to solve it. This approach makes full use of the advantages of the accelerated gradient technique and adaptive parameters updating strategy. Meanwhile, a convergence rate of O(1/N2) can be guaranteed under mild conditions. Experimental results on five standard databases show the superiority of the proposed algorithm over some existing methods.

Deep sparse multi-task learning for feature selection in Alzheimer's disease diagnosis.

Recently, neuroimaging-based Alzheimer's disease (AD) or mild cognitive impairment (MCI) diagnosis has attracted researchers in the field, due to the increasing prevalence of the diseases. Unfortunately, the unfavorable high-dimensional nature of neuroimagingdata, but a limited small number of samples available, makes it challenging to build a robust computer-aided diagnosis system. Machine learning techniques have been considered as a useful tool in this respect and, among various methods, sparse regression has shown its validity in the literature. However, to our best knowledge, the existing sparse regression methods mostly try to select features based on the optimal regression coefficients in one step. We argue that since the training feature vectors are composed of both informative and uninformative or less informative features, the resulting optimal regression coefficients are inevidently affected by the uninformative or less informative features. To this end, we first propose a novel deep architecture to recursively discard uninformative features by performing sparse multi-task learning in a hierarchical fashion. We further hypothesize that the optimal regression coefficients reflect the relative importance of features in representing the target response variables. In this regard, we use the optimal regression coefficients learned in one hierarchy as feature weighting factors in the following hierarchy, and formulate a weighted sparse multi-task learning method. Lastly, we also take into account the distributional characteristics of samples per class and use clustering-induced subclass label vectors as target response values in our sparse regression model. In our experiments on the ADNI cohort, we performed both binary and multi-class classification tasks in AD/MCI diagnosis and showed the superiority of the proposed method by comparing with the state-of-the-art methods.

ODVBA-C: Optimally-Discriminative Voxel-Based Analysis of Continuous Variables.

In this paper, we propose a new method that utilizes a novel spatially adaptive scheme for detection of multivariate neuroimaging patterns relating to a continuous subject-level variable, aiming to effectively determine the optimal spatially adaptive filtering of neuroimaging data from the persepective of finding relationships between imaging and continues (e.g. clinical and cognitive) variables. Analyses employ local pattern analysis using regularized least square regression with nonnegativity constraints within a spatial neighborhood around each voxel. Within each neighborhood, we determine the optimal regression coefficients that relate local patterns to the continuous variable of interest. As each voxel belongs to multiple overlapping neighborhoods, the statistic for a given voxel is determined by combining weights from all neighborhoods to which the voxel participates. Finally, nonparametric permutation testing is used to obtain a voxelwise significance map. Using both simulated and real fMRI data, we demonstrate the effectiveness of the proposed method.

Inversion Analysis of Three-Dimensional Geostress Field in the Second Stage Project of South-to-North Water Transfer

Focus on the topography features of V-shaped deep river valley in Ada dam area, the influence of intensively, moderately and slightly weathered stratum as well as the action of river valley erosion on geostress field was considered. Then, the 3-D initial geostress unloading model was established based on the measured geostress data and geological conditions. Combining the finite element numerical method and inversion theory, multiple regression analysis was established between the measured and corresponding calculated results of geostress. Then, the optimal regression coefficient could be derived. In addition, for convenience of comparison, another method was adopted by ignoring the action of river valley erosion. Through the comparison between the computed values by two methods and the measured geostress values respectively, it is found that the computed values by the former method are more close to the measured ones. Therefore, the distribution rule of initial geostress field can be obtained accurately, which provides important reference for design and construction of Ada dam and its diversion tunnel layout.

Inversion of Initial <i>In Situ</i> Stress Field by the Method of Multivariate Analysis and Engineering Application

Based on engineering geological conditions and measured data of the in-situ stress at a hydropower station, 3D geological models under each affecting factor are calculated by means of finite element computing tools ABAQUS and MATLAB. Then, a multivariate regression model is created between the measured and calculated value of in-situ stress at measurement points, and the optimal regression coefficient of the model is found. Consequently, the distributions of initial in-situ stress of this area are obtained. It is the first time to take into account the independence and internal relations among each stress components. Thus, the more reasonable distributions of initial in-situ stress of this area are obtained. The results indicate that the 3D calculating in-situ stress field is reasonable.