Bug-fix time prediction using non-linear regression through neural network

2021 ◽  
Author(s):  
H. Mahfoodh ◽  
M. Hammad
Keyword(s):  
Neural Network ◽  
Linear Regression ◽  
Time Prediction ◽  
Non Linear

scholarly journals Disparities in particulate matter (PM<sub>10</sub>) origins and oxidative potential at a city-scale (Grenoble, France) – Part II: Sources of PM<sub>10</sub> oxidative potential using multiple linear regression analysis and the predictive applicability of multilayer perceptron neural network analysis

2021 ◽  
Author(s):  
Lucille Joanna S. Borlaza ◽  
Samuël Weber ◽  
Jean-Luc Jaffrezo ◽  
Stephan Houdier ◽  
Rémy Slama ◽  
...  
Keyword(s):  
Neural Network ◽  
Particulate Matter ◽  
Linear Regression ◽  
Multiple Linear Regression ◽  
Multilayer Perceptron ◽  
Mass Concentration ◽  
Redox Activity ◽  
Oxidative Potential ◽  
Atmospheric Processes ◽  
Non Linear

Abstract. The oxidative potential (OP) of particulate matter (PM) quantifies PM capability to cause anti-oxidant imbalance. Due to the wide range and complex mixture of species in particulates, little is known on the pollution sources most strongly contributing to OP. A one-year sampling of PM10 (particles with an aerodynamic diameter below 10) was performed over different sites in a medium-sized city (Grenoble, France). An enhanced fine-scale apportionment of PM10 sources, based on the chemical composition, was performed using Positive Matrix Factorization (PMF) method and reported in a companion paper (Borlaza et al., 2020). OP was assessed as the ability of PM10 to generate reactive oxygen species (ROS) using three different acellular assays: Dithiothreitol (DTT), Ascorbic acid (AA), and 2,7-dichlorofluorescein (DCFH) assays. Using multiple linear regression (MLR), the OP contribution of the sources identified by PMF were estimated. Conversely, since atmospheric processes are usually non-linear in nature, artificial neural network (ANN) techniques, which employs non-linear models, could further improve estimates. Hence, the multilayer perceptron analysis (MLP), an ANN-based model, was additionally used to model OP based on PMF-resolved sources as well. This study presents the spatiotemporal variabilities of OP activity with influences by season-specific sources, site typology and specific local features, and assay sensitivity. Overall, both MLR and MLP effectively captured the evolution of OP. The primary traffic and biomass burning sources were the strongest drivers of OP in the Grenoble basin. There is also a clear redistribution of source-specific impacts when using OP instead of mass concentration, underlining the importance of PM redox activity over mass concentration. Finally, the MLP generally offered improvements in OP prediction especially for sites where synergistic and/or antagonistic effects between sources are prominent, supporting the value of using ANN-based models to account for the non-linear dynamics behind the atmospheric processes affecting OP of PM10.

scholarly journals Comparing the Prediction Capabilities of Artificial Neural Network (ANN) and Nonlinear Regression Models in Pet-Poy Yarn Characteristics and Optimization of Yarn Production Conditions

2017 ◽  
Vol 12 (3) ◽  
pp. 155892501701200 ◽  
Author(s):  
Kenan Yıldirimm ◽  
Hamdi Ogut ◽  
Yusuf Ulcay
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Linear Regression ◽  
Nonlinear Regression ◽  
Fine Tuning ◽  
Draw Force ◽  
Non Linear ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Production Conditions

In the manufacture of yarn, predicting the effect of changing production conditions is vital to reducing defects in the end product. This study compares, for the first time, non-linear regression and artificial neural network (ANN) models in predicting 10 yarn properties shaped by the influence of winding speed, quenching air temperature and/or quenching air speed during production. A multilayer perceptron ANN model was created by training 81 patterns using the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm. The hyperbolic tangent, or TanH, activation function and logistic activation functions were used for the hidden and output layers respectively. Results showed that the ANN approach exhibited a greater prediction capability over the nonlinear regression method. ANN simultaneously predicted all of the 10 final properties of a yarn; tensile strength, tensile strain, draw force, crystallinity ratio, dye uptake based on the colour strengths (K/S), brightness, boiling shrinkage and yarn evenness, more accurately than the non-linear regression model (R2=0.97 vs. R2=0.92). These results lend support to the idea that the ANN analysis combined with optimization can be used successfully to prevent production defects by fine tuning the production environment.

scholarly journals Retrieval of aerosol optical depth from surface solar radiation measurements using machine learning algorithms, non-linear regression and a radiative transfer-based look-up table

2016 ◽  
Vol 16 (13) ◽  
pp. 8181-8191 ◽  
Author(s):  
Jani Huttunen ◽  
Harri Kokkola ◽  
Tero Mielonen ◽  
Mika Esa Juhani Mononen ◽  
Antti Lipponen ◽  
...  
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Support Vector Machine ◽  
Linear Regression ◽  
Support Vector ◽  
Learning Methods ◽  
Surface Solar Radiation ◽  
Machine Learning Methods ◽  
Look Up Table ◽  
Non Linear

Abstract. In order to have a good estimate of the current forcing by anthropogenic aerosols, knowledge on past aerosol levels is needed. Aerosol optical depth (AOD) is a good measure for aerosol loading. However, dedicated measurements of AOD are only available from the 1990s onward. One option to lengthen the AOD time series beyond the 1990s is to retrieve AOD from surface solar radiation (SSR) measurements taken with pyranometers. In this work, we have evaluated several inversion methods designed for this task. We compared a look-up table method based on radiative transfer modelling, a non-linear regression method and four machine learning methods (Gaussian process, neural network, random forest and support vector machine) with AOD observations carried out with a sun photometer at an Aerosol Robotic Network (AERONET) site in Thessaloniki, Greece. Our results show that most of the machine learning methods produce AOD estimates comparable to the look-up table and non-linear regression methods. All of the applied methods produced AOD values that corresponded well to the AERONET observations with the lowest correlation coefficient value being 0.87 for the random forest method. While many of the methods tended to slightly overestimate low AODs and underestimate high AODs, neural network and support vector machine showed overall better correspondence for the whole AOD range. The differences in producing both ends of the AOD range seem to be caused by differences in the aerosol composition. High AODs were in most cases those with high water vapour content which might affect the aerosol single scattering albedo (SSA) through uptake of water into aerosols. Our study indicates that machine learning methods benefit from the fact that they do not constrain the aerosol SSA in the retrieval, whereas the LUT method assumes a constant value for it. This would also mean that machine learning methods could have potential in reproducing AOD from SSR even though SSA would have changed during the observation period.

Fuzzy neural network and coupled gene expression programming/multivariate non-linear regression approach on mechanical features of hydroxyapatite/graphene oxide/epoxy: Empirical and optimization study

2021 ◽  
pp. 095440622110197
Author(s):  
Sasan Fooladpanjeh ◽  
Ali Dadrasi ◽  
Abdorreza Alavi Gharahbagh ◽  
Vali Parvaneh
Keyword(s):  
Neural Network ◽  
Gene Expression ◽  
Mechanical Properties ◽  
Graphene Oxide ◽  
Linear Regression ◽  
Epoxy Resin ◽  
Fuzzy Neural Network ◽  
Gene Expression Programming ◽  
Non Linear ◽  
Fuzzy Neural

One way to enhance the mechanical properties of nanocomposites has been to use different fillers. In this study, ternary hybrid composites of graphene oxide/hydroxyapatite/epoxy resin were investigated. An experimental design was performed based on the central composite design (CCD). Epoxy resin was modified by incorporating different graphene oxide and hydroxyapatite weight from 0 to 0.5 wt.% and 0 to 7 wt.%, respectively. Experimental results showed that Young’s modulus, yield strength and impact strength improved up to 25.64%, 5.95% and 100.05% compared to the neat epoxy resin, respectively. In addition, gene expression programming (GEP), multivariate non-linear regression (MNLR) and fuzzy neural network (FNN) methods were employed to determine the effects of nanoparticles on the mechanical properties. Based on the modelling results, optimization process was investigated by using particle swarm optimization (PSO). Finally, the fracture surface morphologies of the nanocomposites were analyzed by scanning electron microscopy.

Artificial Neural Network Model with Adaptive GMDH Technique for Non-Linear Regression Prediction

2021 ◽  
Vol 12 (5) ◽  
pp. 1142-1150
Author(s):  
Abdul Joseph Fofanah ◽  
Saidu Koroma ◽  
Alpha Bilhal Sesay ◽  
Ahmed Fofana
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Linear Regression ◽  
Prediction Error ◽  
Artificial Neural Network Model ◽  
Theoretical Concept ◽  
Neural Nets ◽  
Model Complexity ◽  
Model Assessment ◽  
Non Linear

A theoretical concept of the GMDH technique using a non-linear regression model, multilayered neural nets, model assessment, and selection to determine the prediction error versus selection model complexity was reviewed and evaluated. The model selection was experimented and evaluated with MATLAB.

scholarly journals Disparities in particulate matter (PM<sub>10</sub>) origins and oxidative potential at a city scale (Grenoble, France) – Part 2: Sources of PM<sub>10</sub> oxidative potential using multiple linear regression analysis and the predictive applicability of multilayer perceptron neural network analysis

2021 ◽  
Vol 21 (12) ◽  
pp. 9719-9739
Author(s):  
Lucille Joanna S. Borlaza ◽  
Samuël Weber ◽  
Jean-Luc Jaffrezo ◽  
Stephan Houdier ◽  
Rémy Slama ◽  
...  
Keyword(s):  
Neural Network ◽  
Particulate Matter ◽  
Linear Regression ◽  
Multiple Linear Regression ◽  
Multilayer Perceptron ◽  
Redox Activity ◽  
Oxidative Potential ◽  
Atmospheric Processes ◽  
Wide Range ◽  
Non Linear

Abstract. The oxidative potential (OP) of particulate matter (PM) measures PM capability to potentially cause anti-oxidant imbalance. Due to the wide range and complex mixture of species in particulates, little is known about the pollution sources most strongly contributing to OP. A 1-year sampling of PM10 (particles with an aerodynamic diameter below 10) was performed over different sites in a medium-sized city (Grenoble, France). An enhanced fine-scale apportionment of PM10 sources, based on the chemical composition, was performed using the positive matrix factorization (PMF) method and reported in a companion paper (Borlaza et al., 2020). OP was assessed as the ability of PM10 to generate reactive oxygen species (ROS) using three different acellular assays: dithiothreitol (DTT), ascorbic acid (AA), and 2,7-dichlorofluorescein (DCFH) assays. Using multiple linear regression (MLR), the OP contributions of the sources identified by PMF were estimated. Conversely, since atmospheric processes are usually non-linear in nature, artificial neural network (ANN) techniques, which employ non-linear models, could further improve estimates. Hence, the multilayer perceptron analysis (MLP), an ANN-based model, was additionally used to model OP based on PMF-resolved sources as well. This study presents the spatiotemporal variabilities of OP activity with influences by season-specific sources, site typology and specific local features, and assay sensitivity. Overall, both MLR and MLP effectively captured the evolution of OP. The primary traffic and biomass burning sources were the strongest drivers of OP in the Grenoble basin. There is also a clear redistribution of source-specific impacts when using OP instead of mass concentration, underlining the importance of PM redox activity for the identification of potential sources of PM toxicity. Finally, the MLP generally offered improvements in OP prediction, especially for sites where synergistic and/or antagonistic effects between sources are prominent, supporting the value of using ANN-based models to account for the non-linear dynamics behind the atmospheric processes affecting OP of PM10.

Sign in / Sign up

Export Citation Format

Share Document