Among those predicted parameters, Vd was the most precise one that 95% predicted value was in the 95% confidence limits for "observed" value with the correlation coefficient between predicted Vd value and the "observed" value being 0.974, the estimated standard error (SE) was less than 3. 3%;

The calculation results show that the statistic and monitoring value are fairly identical, multiple correlation coefficients bigger and estimated standard error smaller.

The predicted CL that 2/3 predicted value were in the 95% confidence limits for the "observed" value was less precise with the correlation coefficient being 0. 797 and the SE less than 20%;

Thirdly, a photo-thermal based crop development and growth model system was developed for prediction of crop developmental stage, biomass production and yield.

模型对夏干季节(三伏天)、夏湿季节(梅雨季节)和冬季温室内空气温度、湿度以及作物蒸腾速率的预测结果与1：1直线之间的决定系数R~2和回归估计标准误差RMSE(Root Mean SquaredError)分别为：0.89，0.75，0.52；

Therefore, when two or three smoothing coefficients given at random are used to compare standard errors, what is selected from them is not the best value, so best estimating results are not obtained. In this paper, through analysis and study, a practical and convenient method is given to solve these two problems.

The root mean squared error (RMSE) for germination, seedling, flowering, fruit setting, and harvest development stages was 0, 1, 1.87, 2.69 and 3 days, respectively. The prediction accuracy of this model is remarkably higher than that of the Growth Degree Day (GDD) based model (RMSE is 0, 7.91, 8.86, 13.58 and 12.59 days, respectively, for germination, seedling, flowering, fruit setting, and harvest development stages).

The results show that the simulated results agree well with the observed ones. The root mean squared error (RMSE) for development stages of germination, five leaf stage, tendril elongation, flowering, fruit setting and the whole growth duration was 1, 3, 2.2, 1.8, 1.1, 2.6 days, respectively.

These data allow an in vivo prediction of vertebral body strength using a noninvasive method with a standard error of estimate amounting to less than 0.95 kN.

The best formula was: Expected percent dose excretion at 35 min = 79.3[1-e-(0.004798 x ERPF)] with a standard error of estimate (Sy·x) of 5.2% dose.

gave the lowest standard error of estimate (Sy·x) of all the methods.

Standard error of estimate, bias and imprecision of different methods were evaluated.

On an hourly basis, the overall standard error of estimate (SEE) and the absolute relative error (ARE) were 0.06?mm h-1 (41?W m-2) and 4.2%, respectively.

This expression applies to the range 100-1 000 MeV with an estimated standard error of ±20-25%.

When only the number,Nmax of tracks in the compartment in which the shower has its maximum development can be determined, our best estimate isE0=87.4Nmax, with an estimated standard error of ±30-35%.

Based on the criterion of estimated standard error, the efficiency of the improved estimator with respect to the traditional unbiased estimator (i.e., sample mean) is examined numerically.

The estimated standard error in μO2ss is on the order of ±200 J mol-1, which is approximately ±0.01 log-bar units in fo2 at 1273 K.

When estimating the mean value of a variable, or the total amount of a resource, within a specified region it is desirable to report an estimated standard error for the resulting estimate.

If the sample sites are selected according to a probability sampling design, it usually is possible to construct an appropriate design-based standard error estimate.

Hence, using one standard error estimate for all genes does not seem appropriate.

Of these, the EM covariance matrix with average column-wise n gave estimates closest to the true values with the closest standard error estimate.

The coefficient of multiple determination and standard error estimate for Equation 7 were found to be 0.99 and 0.067, respectively.

Using the method of one point of Bayesian individual pharmacokinetic parameters and dosage regimens of digoxin were calculated in heart failure patients,and the population pharmacokinetic data analysed in statistics with a minitab software. While CrCL being steady-reaction variable,the series of regression models were suggested and the individual pharmacokinetic parameters of digoxin in heart failure patients,individual dosage regimen as well as the corresponding average steady state serum concencration (Css)...

Using the method of one point of Bayesian individual pharmacokinetic parameters and dosage regimens of digoxin were calculated in heart failure patients,and the population pharmacokinetic data analysed in statistics with a minitab software. While CrCL being steady-reaction variable,the series of regression models were suggested and the individual pharmacokinetic parameters of digoxin in heart failure patients,individual dosage regimen as well as the corresponding average steady state serum concencration (Css) were predicted. The results showed that there were strong correlations (P<0. 001) between CrCL and the population pharmacokinetic parameters of digoxin,and the correlation coefficients were CL = 0. 805,Vd=0. 985,Tl/2 = 0. 517,K = 0. 525,D1.0=0. 714 respectively. The individual pharmacokinetic parameters of digoxin predicted by the founding of quantity regression models were not significant different to the "observed" parameters . Among those predicted parameters, Vd was the most precise one that 95% predicted value was in the 95% confidence limits for "observed" value with the correlation coefficient between predicted Vd value and the "observed" value being 0.974, the estimated standard error (SE) was less than 3. 3%; The predicted CL that 2/3 predicted value were in the 95% confidence limits for the "observed" value was less precise with the correlation coefficient being 0. 797 and the SE less than 20%; The predicted T1/2 and K were a little bad. 80% dosage regiments that were predicted according to CrCL corresponds with the best individual dosage regiments and the corresponding Css of either wasn't significent difference.

There are two problems to be solved if index smoothing method is directly used for prediction: determination of initial value and smoothing coefficient, both of which not only directly affect predicting results, but also have internal associations. Therefore, when two or three smoothing coefficients given at random are used to compare standard errors, what is selected from them is not the best value, so best estimating results are not obtained. In this paper, through analysis and study, a practical and convenient...

There are two problems to be solved if index smoothing method is directly used for prediction: determination of initial value and smoothing coefficient, both of which not only directly affect predicting results, but also have internal associations. Therefore, when two or three smoothing coefficients given at random are used to compare standard errors, what is selected from them is not the best value, so best estimating results are not obtained. In this paper, through analysis and study, a practical and convenient method is given to solve these two problems.

One variable linear regressive equation is an anticipating method which combines the regressive analysis in statistics with anticipation theory.It's very practical.Firstly,we should see if there exist linear relations among the economic variations.Then,we should get the regressive equation to anticipate parameters with the method of OLS.Finally,we must calculate the standard estimated error to make sure the degree of regressive model's credibility.