A Comparative Analysis on Prediction Performance of Regression Models during Machining of Composite Materials

Modeling the interrelationships between the input parameters and outputs (responses) in any machining processes is essential to understand the process behavior and material removal mechanism. The developed models can also act as effective prediction tools in envisaging the tentative values of the responses for given sets of input parameters. In this paper, the application potentialities of nine different regression models, such as linear regression (LR), polynomial regression (PR), support vector regression (SVR), principal component regression (PCR), quantile regression, median regression, ridge regression, lasso regression and elastic net regression are explored in accurately predicting response values during turning and drilling operations of composite materials. Their prediction performance is also contrasted using four statistical metrics, i.e., mean absolute percentage error, root mean squared percentage error, root mean squared logarithmic error and root relative squared error. Based on the lower values of those metrics and Friedman rank and aligned rank tests, SVR emerges out as the best performing model, whereas the prediction performance of median regression is worst. The results of the Wilcoxon test based on the drilling dataset identify the existence of statistically significant differences between the performances of LR and PCR, and PR and median regression models.

Correlates of sugar-sweetened beverage intake among low-income women during the first 1000 days

Objective: To describe prenatal and postpartum consumption of water, cows’ milk, 100 % juice and sugar-sweetened beverages (SSB) among women enrolled in the Special Supplemental Nutrition Program for Women, Infants and Children (WIC) programme in New York City (NYC) and to identify correlates of SSB intake in this population. Design: Cross-sectional data were collected from structured questionnaires that included validated beverage frequency questionnaires with the assistance of container samples. The association of maternal and household factors and non-SSB consumption with habitual daily energetic (kJ (kcal)) intake from SSB was assessed by using multivariable median regression. Setting: WIC programme in NYC, NY. Data were collected in 2017. Participants: 388 pregnant or postpartum women (infant aged <2 years) from the NYC First 1000 Days Study. Results: Median age was 28 years (interquartile range (IQR) 24–34); 94·1 % were Hispanic/Latina, and 31·4 % were pregnant. Overall, 87·7 % of pregnant and 89·1% of postpartum women consumed SSB ≥ once weekly, contributing to a median daily energetic intake of 410 kJ (98 kcal) (IQR (113–904 kJ) 27–216) and 464 kJ (111 kcal) (IQR (163–1013 kJ) 39–242), respectively. In adjusted analyses, only consumption of 100 % juice was associated with greater median energetic intake from SSB (adjusted β for each additional ounce = 13; 95% CI 8, 31 (3·2; 95 % CI 2·0, 7·3). Conclusions: Among pregnant and postpartum women in WIC-enrolled families, interventions to reduce SSB consumption should include reduction of 100 % juice consumption as a co-target of the intervention.

Does Culture Affect Tourism Demand? A Global Perspective

Tourism studies commonly focus on the determinants of tourism demand. While most examine factors such as economic determinants, research on the effect of culture on tourism demand remains underdeveloped. This study uses a Bayesian two-stage median regression method to eliminate the potential collinearity between cultural and travel distance and to estimate the impact of cultural distance more appropriately. The results show that while there is a negative relationship between cultural distance and tourism demand, tourism demand is less sensitive to change in cultural distance; the popularity of a travel route moderates the effect of cultural distance on tourism demand; and the influence of cultural distance is different across time and different source markets.

Avaliação relativa de desempenho na compensação pecuniária de executivos: evidências empíricas no Brasil

The objective of this study was to analyze the implicit use of relative performance evaluation in BM&FBovespa listed companies as a way to measure the remuneration of its executives. To define the sample, we sought to identify companies that disclosed information about the compensation of their executives between 2009 and 2012, totaling the sample size in 67 companies, totaling 112 observations. They were then categorized in order to capture risk sharing as predicted by the theory of relative performance evaluation. The results of this research indicate a strong asymmetry in the distribution of the compensation, mainly due to the long-term compensation, which caused the occurrence of outliers. As a result of this situation, and following studies already developed, it was decided to test the model through quantile regression. Even with the use of the median regression it was not possible to identify statistically significant evidences of the occurrence of relative performance evaluation, therefore, there is no evidence that the variation of the result of the sector reduces the impacts that the results obtained by the organizations exercise on the executive remuneration.

Regresi Median Pada Copula Bivariat

Abstrak: Analisis regresi adalah analisis yang sering digunakan dalam segala bidang yang bertujuan untuk memodelkan hubungan antara dua jenis variabel tak bebas dengan satu atau variabel bebas. Regresi linier masih memiliki beberapa kekurangan, maka dari untuk mengatasinya dengan regresi median. Copula dapat digunakan untuk mendeteksi hubungan data bivariat dengan peubah-peubah yang berbeda. Hasil penelitian menunjukkan kurva kuantil bersyarat terbaik berdasarkan MSE terkecil Data I yaitu copula Plackett sebesar 0.8650. Sedangkan nilai MSE terkecil Data II yaitu copula Gaussian sebesar 0.3954. Nilai MSE terkecil Data III yaitu copula Frank sebesar 0.5575. Terakhir, nilai MSE terkecil Data IV yaitu copula Clayton sebesar 0.3190.Abstract: Regression analysis is an analysis that is often used in all fields which aims to model the relationship between two types of non-dependent variables with one or independent variables. Linear regression still has several drawbacks, so to overcome this by median regression. Copula can be used to detect bivariate data relations with different variables. The results showed that the best conditional curves based on the smallest MSE of Data I were Plackett copula of 0.8650. While the smallest MSE value is Data II, which is a Gaussian population of 0.3954. The smallest MSE value of Data III is Frank copula of 0.5575. Finally, the smallest MSE value is Data IV which is copula Clayton of 0.3190.