Ridge Regression Based Development of Norris-Landzberg Acceleration Factors and Goldmann Constants for Leadfree Electronics

2011 ◽  
Author(s):  
Pradeep Lall ◽  
Dinesh Arunachalam ◽  
Jeff Suhling
Keyword(s):  
Ridge Regression ◽  
Thermal Cycle ◽  
Regression Models ◽  
Coefficient Of Thermal Expansion ◽  
Model Development ◽  
High Sensitivity ◽  
Maximum Temperature ◽  
Temperature Cycle ◽  
Linear Regression Models ◽  
Part Selection

Goldmann Constants and Norris-Landzberg acceleration factors for lead-free solders have been developed based on ridge regression models (RR) for reliability prediction and part selection of area-array packaging architectures under thermo-mechanical loads. Ridge regression adds a small positive bias to the diagonal of the covariance matrix to prevent high sensitivity to variables that are correlated. The proposed procedure proves to be a better tool for prediction than multiple-linear regression models. Models have been developed in conjunction with Stepwise Regression Methods for identification of the main effects. Package architectures studied include, BGA packages mounted on copper-core and no-core printed circuit assemblies in harsh environments. The models have been developed based on thermo-mechanical reliability data acquired on copper-core and no-core assemblies in four different thermal cycling conditions. Packages with Sn3Ag0.5Cu solder alloy interconnects have been examined. The models have been developed based on perturbation of accelerated test thermo-mechanical failure data. Data has been gathered on nine different thermal cycle conditions with SAC305 alloys. The thermal cycle conditions differ in temperature range, dwell times, maximum temperature and minimum temperature to enable development of constants needed for the life prediction and assessment of acceleration factors. Norris-Landzberg acceleration factors have been benchmarked against previously published values. In addition, model predictions have been validated against validation datasets which have not been used for model development. Convergence of statistical models with experimental data has been demonstrated using a single factor design of experiment study for individual factors including temperature cycle magnitude, relative coefficient of thermal expansion, and diagonal length of the chip. The predicted and measured acceleration factors have also been computed and correlated. Good correlations have been achieved for parameters examined.

PCR for Derivation of Parameter Dependencies, Thermo-Mechanical Norris-Landzberg Acceleration Factors, Goldmann Fatigue Constants for Leadfree Electronics

2009 ◽  
Author(s):  
Pradeep Lall ◽  
Aniket Shirgaokar ◽  
Dineshkumar Arunachalam ◽  
Jeff Suhling ◽  
Mark Strickland ◽  
...  
Keyword(s):  
Test Data ◽  
Thermal Cycle ◽  
Regression Models ◽  
Reliability Prediction ◽  
Maximum Temperature ◽  
Temperature Cycle ◽  
Data Sets ◽  
Accelerated Test ◽  
Bga Packages ◽  
Part Selection

Goldmann Constants and Norris-Landzberg acceleration factors for lead-free solders have been developed based on principal component regression models (PCR) for reliability prediction and part selection of area-array packaging architectures under thermo-mechanical loads. Models have been developed in conjunction with Stepwise Regression Methods for identification of the main effects. Package architectures studied include, BGA packages mounted on copper-core and no-core printed circuit assemblies in harsh environments. The models have been developed based on thermo-mechanical reliability data acquired on copper-core and no-core assemblies in four different thermal cycling conditions. Packages with Sn3Ag0.5Cu solder alloy interconnects have been examined. The models have been developed based on perturbation of accelerated test thermo-mechanical failure data. Data has been gathered on nine different thermal cycle conditions with SAC305 alloys. The thermal cycle conditions differ in temperature range, dwell times, maximum temperature and minimum temperature to enable development of constants needed for the life prediction and assessment of acceleration factors. Goldmann Constants and the Norris-Landzberg acceleration factors have been benchmarked against previously published values. In addition, model predictions have been validated against validation data-sets which have not been used for model development. Convergence of statistical models with experimental data has been demonstrated using a single factor design of experiment study for individual factors including temperature cycle magnitude, relative coefficient of thermal expansion, and diagonal length of the chip. The predicted and measured acceleration factors have also been computed and correlated. Good correlations have been achieved for parameters examined. Previously, the feasibility of using multiple linear regression models for reliability prediction has been demonstrated for flex-substrate BGA packages [Lall 2004, 2005], flip-chip packages [Lall 2005] and ceramic BGA packages [Lall 2007]. The presented methodology is valuable in the development of fatigue damage constants for the application specific accelerated test data-sets and provides a method to develop institutional learning based on prior accelerated test data.

scholarly journals Inflammatory Markers Are Positively Associated with Serumtrans-Fatty Acids in an Adult American Population

2017 ◽  
Vol 2017 ◽  
pp. 1-6 ◽  
Author(s):  
Mohsen Mazidi ◽  
Hong-kai Gao ◽  
Andre Pascal Kengne
Keyword(s):  
Fatty Acids ◽  
Linear Regression ◽  
Regression Models ◽  
Survey Design ◽  
High Sensitivity ◽  
Analysis Of Covariance ◽  
Linear Regression Models ◽  
Inflammatory Profile ◽  
Hs Crp ◽  
High Level

Background and Aim. The relationship between serumtrans-fatty acids (TFAs) and systemic inflammation markers is unclear. We investigated the association of serum TFAs with high sensitivity C-reactive protein (hs-CRP) and fibrinogen in adult Americans.Methods. The 1999 to 2000 National Health and Nutrition Examination Survey (NHANES) participants with measured data on hs-CRP and fibrinogen were included. TFAs were measured via capillary gas chromatography and mass spectrometry using negative chemical ionization. Analysis of covariance and multivariable-adjusted linear regression models were used to investigate the associations between these parameters, accounting for the survey design.Results. Of the 5446 eligible participants, 46.8% (n=2550) were men. The mean age was 47.1 years overall: 47.8 years in men and 46.5 years in women (p=0.085). After adjustment for age and sex, mean serum TFAs rose with the increasing quarters of hs-CRP and fibrinogen (bothp<0.001). In linear regression models adjusted for age, sex, race, education, marital status, body mass index, and smoking, serum TFAs were an independent predictor of plasma hs-CRP and fibrinogen levels.Conclusion. A high level of TFAs appears to be a contributor to an unfavourable inflammatory profile. Because serum TFAs concentrations are affected by dietary TFA intake, these data suggest a possible contribution of TFAs intake modulation in the prevention of inflammation-related chronic diseases.

Catch-–effort maximum likelihood estimation of important population parameters

1997 ◽  
Vol 54 (4) ◽  
pp. 890-897 ◽  
Author(s):  
W R Gould ◽  
K H Pollock
Keyword(s):  
Maximum Likelihood ◽  
Maximum Likelihood Estimation ◽  
Regression Models ◽  
Model Development ◽  
Likelihood Estimation ◽  
Maximum Likelihood Estimates ◽  
Linear Regression Models ◽  
Estimation Of Parameters ◽  
Regression Methods ◽  
Regression Techniques

The relative ease with which linear regression models are understood explains the popularity of such techniques in estimating population size with catch-effort data. However, the development and use of the regression models require assumptions and approximations that may not accurately reflect reality. We present the model development necessary for maximum likelihood estimation of parameters from catch-effort data using the program SURVIV, the primary intent being to present biologists with a vehicle for producing maximum likelihood estimates in lieu of using the traditional regression techniques. The differences between the regression approaches and maximum likelihood estimation will be illustrated with an example of commercial fishery catch-effort data and through simulation. Our results indicate that maximum likelihood estimation consistently provides less biased and more precise estimates than the regression methods and allows for greater model flexibility necessary in many circumstances. We recommend the use of maximum likelihood estimation in future catch-effort studies.

scholarly journals Model development for the estimation of urban air temperature based on surface temperature and NDVI

2020 ◽  
Vol 54 (3) ◽  
pp. 29-40 ◽  
Author(s):  
Y Guo ◽  
Tamás Gál ◽  
Guohang Tian ◽  
János Unger
Keyword(s):  
Linear Regression ◽  
Surface Temperature ◽  
Air Temperature ◽  
Land Surface ◽  
Regression Models ◽  
Pearson Correlation ◽  
Model Development ◽  
Urban Air ◽  
Landsat 8 ◽  
Linear Regression Models

Predictive models for urban air temperature (Tair) were developed by using urban land surface temperature (LST) retrieved from Landsat-8 and MODIS data, NDVI retrieved from Landsat-8 data and Tair measured by 24 climatological stations in Szeged. The investigation focused on summer period (June−September) during 2016−2019 in Szeged. The relationship between Tair and LST was analyzed by calculating Pearson correlation coefficient, root-mean-square error and mean-absolute error using the data of 2017−2019, then unary (LST) and binary (LST and NDVI) linear regression models were developed for estimating Tair. The data in 2016 were used to validate the accuracy of the models. Correlation analysis indicated that there were strong correlations during the nighttime and relatively weaker ones during the daytime. The errors between Tair and LSTMODIS-Night was the smallest, followed by LSTMODIS-Day and LSTLandsat-8 respectively. The validation results showed that all models could perform well, especially during nighttime with an error of less than 1.5o. However, the addition of NDVI into the linear regression models did not significantly improve the accuracy of the models, and even had a negative effect. Finally, the influencing factors and temporal and spatial variability of the correlation between Tair and LST were analyzed. LSTLandsat-8 had a larger original error with Tair, but the regression model based on Landsat-8 had a stronger ability to reduce errors.

scholarly journals True versus perturbed forest inventory plot locations for modeling: a simulation study

2006 ◽  
Vol 36 (3) ◽  
pp. 801-807 ◽  
Author(s):  
John W Coulston ◽  
Kurt H Riitters ◽  
Ronald E McRoberts ◽  
Greg A Reams ◽  
William D Smith
Keyword(s):  
Linear Regression ◽  
Mean Square Error ◽  
Forest Inventory ◽  
Regression Models ◽  
Model Development ◽  
Forest Health ◽  
Mean Square ◽  
Linear Regression Models ◽  
Coarse Spatial Resolution ◽  
The Difference

USDA Forest Service Forest Inventory and Analysis plot information is widely used for timber inventories, forest health assessments, and environmental risk analyses. With few exceptions, true plot locations are not revealed; the plot coordinates are manipulated to obscure the location of field plots and thereby preserve plot integrity. The influence of perturbed plot locations on the development and accuracy of statistical models is unknown. We tested the hypothesis that the influence is related to the spatial structure of the data used in the models. For ordinary kriging we examined the difference in mean square error based on true and perturbed plot locations across a range of spatial autocorrelations. We also examined the difference in mean square error for regression models developed with true and perturbed plot locations across a range of spatial autocorrelations and spatial resolutions. Perturbing plot locations did not significantly influence the accuracy of kriging estimates, but in some situations linear regression model development and accuracy were significantly influenced. Unless the independent variable has high spatial autocorrelation, only coarse spatial resolution data should be used to develop linear regression models.

scholarly journals The Comparison Between Different Approaches to Overcome the Multicollinearity Problem in Linear Regression Models

2018 ◽  
Vol 31 (1) ◽  
pp. 212
Author(s):  
Hazim Mansoor Gorgees ◽  
Fatimah Assim Mahdi
Keyword(s):  
Ridge Regression ◽  
Regression Models ◽  
Estimation Method ◽  
Principal Component ◽  
Ordinary Least Squares ◽  
Linear Regression Models ◽  
Principal Component Method ◽  
Regression Estimators ◽  
The Mean ◽  
Collinearity Problem

    In the presence of multi-collinearity problem, the parameter estimation method based on the ordinary least squares procedure is unsatisfactory. In 1970, Hoerl and Kennard insert analternative method labeled as estimator of ridge regression. In such estimator, ridge parameter plays an important role in estimation. Various methods were proposed by many statisticians to select the biasing constant (ridge parameter). Another popular method that is used to deal with the multi-collinearity problem is the principal component method. In this paper,we employ the simulation technique to compare the performance of principal component estimator with some types of ordinary ridge regression estimators based on the value of the biasing constant (ridge parameter). The mean square error (MSE) is used as a criterion to assess the performance of such estimators.

Sign in / Sign up

Export Citation Format

Share Document