Medical research data are often skewed and heteroscedastic. It has therefore become practice to log-transform data in regression analysis, in order to stabilize the variance. Regression analysis on log-transformed dat...Medical research data are often skewed and heteroscedastic. It has therefore become practice to log-transform data in regression analysis, in order to stabilize the variance. Regression analysis on log-transformed data estimates the relative effect, whereas it is often the absolute effect of a predictor that is of interest. We propose a maximum likelihood (ML)-based approach to estimate a linear regression model on log-normal, heteroscedastic data. The new method was evaluated with a large simulation study. Log-normal observations were generated according to the simulation models and parameters were estimated using the new ML method, ordinary least-squares regression (LS) and weighed least-squares regression (WLS). All three methods produced unbiased estimates of parameters and expected response, and ML and WLS yielded smaller standard errors than LS. The approximate normality of the Wald statistic, used for tests of the ML estimates, in most situations produced correct type I error risk. Only ML and WLS produced correct confidence intervals for the estimated expected value. ML had the highest power for tests regarding β1.展开更多
The purpose of this study was to determine a suitable model for investigating the effects of climate factors on the area burned by forest fire in the Tahe forest region, Daxing＇an Mountains, in northeast China. The r...The purpose of this study was to determine a suitable model for investigating the effects of climate factors on the area burned by forest fire in the Tahe forest region, Daxing＇an Mountains, in northeast China. The response variables were the area burned by lightning- caused fire, human-caused fire, and total burned area. The predictor variables were nine climate variables collected from the local weather station. Three regression models were utilized, including multiple linear regression, log- linear model （log-transformation on both response and predictor variables）, and gamma-generalized linear model. The goodness-of-fit of the models were compared based on model fitting statistics such as R2, AIC, and RMSE. The results revealed that the gamma-generalized linear model was generally superior to both multiple linear regressionmodel and log-linear model for fitting the fire data. Further, the best models were selected based on the criteria that the climate variables were statistically significant at at = 0.05. The gamma best models indicated that maximum wind speed, precipitation, and days that rainfall greater than 0.1 mm had significant impacts on the area burned by the lightning-caused fire, while the mean temperature and minimum relative humidity were the .main drivers of the burned area caused by human activities. Overall, the total burned area by forest fire was significantly influenced by days that rainfall greater than 0.1 mm and minimum rela- tive humidity, indicating that the moisture condition of forest stands determine the burned area by forest fire.展开更多
Laboratory cased-hole acoustic logging simulations are developed with the linear phased-array transmitter in scaled cased well models for evaluating the feasibility of extracting formation acoustic parameters through ...Laboratory cased-hole acoustic logging simulations are developed with the linear phased-array transmitter in scaled cased well models for evaluating the feasibility of extracting formation acoustic parameters through casing.The full waveforms are measured with different cement bonding models.By analyzing the measured wavetrains and the time-slowness correlation graphs,it is showed that when the generation conditions of the refracted compressional wave and the refracted shear wave are reached successively by regulating the direction of acoustic beam radiated from the linear phased-array transmitter,steered angle of the main radiation lobe with both good bonding interfaces.The refracted compressional wave and the refracted shear wave can be stimulated obviously and the casing wave can be suppressed effectively,even when the casing and cement（or the cement and formation） is not bonded.Based on these observations, it is worthwhile to apply the linear phased-array transmitter to determine formation velocities,particularly in poorly bonded cased well.The works establish the experimental and theoretical foundation for new generation cased-hole acoustic logging tool development.展开更多
文摘Medical research data are often skewed and heteroscedastic. It has therefore become practice to log-transform data in regression analysis, in order to stabilize the variance. Regression analysis on log-transformed data estimates the relative effect, whereas it is often the absolute effect of a predictor that is of interest. We propose a maximum likelihood (ML)-based approach to estimate a linear regression model on log-normal, heteroscedastic data. The new method was evaluated with a large simulation study. Log-normal observations were generated according to the simulation models and parameters were estimated using the new ML method, ordinary least-squares regression (LS) and weighed least-squares regression (WLS). All three methods produced unbiased estimates of parameters and expected response, and ML and WLS yielded smaller standard errors than LS. The approximate normality of the Wald statistic, used for tests of the ML estimates, in most situations produced correct type I error risk. Only ML and WLS produced correct confidence intervals for the estimated expected value. ML had the highest power for tests regarding β1.
基金funded by Asia-Pacific Forests Net(APFNET/2010/FPF/001)National Natural Science Foundation of China(Grant No.31400552)Forestry industry research special funds for public welfare projects(201404402)
文摘The purpose of this study was to determine a suitable model for investigating the effects of climate factors on the area burned by forest fire in the Tahe forest region, Daxing＇an Mountains, in northeast China. The response variables were the area burned by lightning- caused fire, human-caused fire, and total burned area. The predictor variables were nine climate variables collected from the local weather station. Three regression models were utilized, including multiple linear regression, log- linear model （log-transformation on both response and predictor variables）, and gamma-generalized linear model. The goodness-of-fit of the models were compared based on model fitting statistics such as R2, AIC, and RMSE. The results revealed that the gamma-generalized linear model was generally superior to both multiple linear regressionmodel and log-linear model for fitting the fire data. Further, the best models were selected based on the criteria that the climate variables were statistically significant at at = 0.05. The gamma best models indicated that maximum wind speed, precipitation, and days that rainfall greater than 0.1 mm had significant impacts on the area burned by the lightning-caused fire, while the mean temperature and minimum relative humidity were the .main drivers of the burned area caused by human activities. Overall, the total burned area by forest fire was significantly influenced by days that rainfall greater than 0.1 mm and minimum rela- tive humidity, indicating that the moisture condition of forest stands determine the burned area by forest fire.
基金supported by the National Natural Science Foundation of China(40804020)Natural Science Foundation of Shandong(ZR2011DQ020)
文摘Laboratory cased-hole acoustic logging simulations are developed with the linear phased-array transmitter in scaled cased well models for evaluating the feasibility of extracting formation acoustic parameters through casing.The full waveforms are measured with different cement bonding models.By analyzing the measured wavetrains and the time-slowness correlation graphs,it is showed that when the generation conditions of the refracted compressional wave and the refracted shear wave are reached successively by regulating the direction of acoustic beam radiated from the linear phased-array transmitter,steered angle of the main radiation lobe with both good bonding interfaces.The refracted compressional wave and the refracted shear wave can be stimulated obviously and the casing wave can be suppressed effectively,even when the casing and cement（or the cement and formation） is not bonded.Based on these observations, it is worthwhile to apply the linear phased-array transmitter to determine formation velocities,particularly in poorly bonded cased well.The works establish the experimental and theoretical foundation for new generation cased-hole acoustic logging tool development.
