绝对误差

实验结果表明，本算法较之传统推荐算法和SlopeOne算法在平均绝对误差值上有一定的提高，证明了本算法的可行性与有效性。

Experimental results show that the algorithm is better in the mean absolute error against traditional algorithm and the Slope One algorithm .

利用时间乘以绝对误差积分准则，设计了直线一级倒立摆的PID控制器。

By using the integral criterion of time multiplies the absolute error , we design the linear 1-stage inverted pendulum PID controller .

研究表明采用最小均方误差或绝对误差准则的块匹配运动估值算法对H。

This paper proposes a method to improve H.261 video coder by modifying the block matching motion estimation algorithm .

为此，在小波去噪的基础上，提出了一种最小绝对误差（LA）辨识算法。

Hence , An identification algorithm based on wavelet and least absolute errors ( WLA ) is proposed .

经仿真分析，SVM模型90%以上预测数据的绝对误差控制在0.01MPa；

According to simulating analysis , the absolute error of 90 % forecasting data is smaller than 0 . 01 MPa .

结果：预测值与实测值的平均绝对误差E为0.532，平均相对误差为1.059%。

Results : Factual value and predicting value and average and absolute error margin E is 0.532 , the average and opposite error margin is 1.059 % .

相关评价指标和绝对误差指标联合应用对QRS波群分辨率高达99.48%。

The resolution rate of QRS complex was 99.48 % both with relative index and absolute error index .

结果表明：温度残差的平均绝对误差（MAE）是Spline平均相对误差（RME）分别为：9.56%和9.66%。

The results of mean absolute errors were ranked as Spline relative mean errors ( RME ) averaged 9.56 % .

用待测波形与模板的绝对误差和相对近似值作为QRS波群的评价指标。

Absolute error and relative approximation were used as estimated indexes of semblance between testing QRS complex and sinus heart rhythm model .

Cl、S的相对标准偏差分别为0.22%和0.36%，最大绝对误差分别为0.43%和0.80%。

The relative standard deviation of chlorine and sulfur are 0.22 % and 0.36 % respectively , the maximal absolute deviation are 0.43 % and 0.80 % .

室内实验中锑电极测定两种标准缓冲液的pH值，当温度小于60℃时，最大绝对误差为0.23；温度为70℃时为0.46。

The result of stibium electrode 's determining pH values of two standard buffer fluid in laboratory showed that the maximum absolute error was 0.23 when temperature went below 60 ℃ and 0.46 when 70 ℃ .

模拟结果与BP网络方法、RBF网络方法相比，在绝对误差、相对误差、相关系数以及数据方差方面都是较优的。

In comparison with BP method and RBF method , the SVM method is proved to be the best in terms of absolute error 、 relative error 、 correlation coefficient and variance .

在顶层进行匹配误差计算时采用修正累计绝对误差（SAD）准则，平滑了矢量场，并进一步提高了运动估计的准确性。

A modified summed absolute difference ( SAD ) criterion was used for motion estimation of the top level , which could smooth the motion vector field and resulted in an improvement of estimation accuracy .

实例分析表明，提出的WNN网络的评价绝对误差平均为0.425%，而BP网络评价绝对误差平均为3.1%。

It is exemplified that the mean absolute error of the estimation results via WNN network as proposed is 0.425 % , while that via BP network is up to 3.1 % .

对比结果表明，两者剥蚀厚度的估算结果相近，绝对误差最大不超过50m，相对误差在20%以内。

The results show that these methods gain the similar denudation thickness with absolute error less than 50m , and relative error limited in 20 % .

对应每一个样本参数仿真计算控制器的阶跃响应，并按时间乘绝对误差积分（integraloftimemultipliedbytheabsolutevalueoferror，ITAE）准则计算目标函数值；

Next , the step response of controller is simulated for each pair of the sample parameters , and the values of objective function are calculated with ITAE ( integral of time multiplied by the absolute value of error ) performance index .

与仅以氨基酸组成为特征矢量的BP神经网络方法比较，相应的他检验平均绝对误差分别减小了0.024和0.016，标准偏差分别减小了0.031和0.018；

Compared with the BP neural network method only using the amino acid composition as features , the average absolute errors for cross-validation test are relatively reduced by 0.024 and 0.016 with the standard deviations reduced by 0.031 and 0.018 , respectively .

因单一预测模型本身具有局限性，然后引入最小误差平方和组合预测模型、最小绝对误差和组合预测模型及BP神经网络组合预测模型这三个组合预测模型进一步进行矿产品价格分析预测。

Due to the limitation of Single forecasting model , the minimum sum of squared errors combined forecasting model , the minimum absolute errors combined forecasting model and Back-Propagation neural network ensemble forecasting model were imported for further analysis and forecast of mineral prices .

在此基础上采用单纯形法对PID参数进行基于时间乘绝对误差积分（ITAE）指标的最优整定。

Then an optimized tuning of PID parameter was carried out using optimized method of many variables with no restriction based on the target of integral of time multiplied absolute error ( ITAE ) .

同时，论文利用绝对误差和相对误差的计算方式，在模拟图像上，比较了本文算法采用的L1模度量和最小二乘算法，分析表明采用的L1模度量算法误差较小。

Then , on the simulated images the thesis uses absolute error and relative error calculation method . Compared with the least squares algorithm , the presented using L1 norm measure has small error .

在最小平均绝对误差（MAE）准则下，实现了复合顺序形态滤波的结构元素自适应处理。

In this paper , the adaptive processing of the structuring element in complex ranked-order morphological filtering is implemented under the criterion of minimizing the mean absolute error ( MAE ) .

该方法具有需要数据少、预测速度快和准确度高的特点，实际应用表明，深度预测的绝对误差小于2m，可用于现场钻井取心卡准层位的精确预报。

The practical application indicates that its absolute error in depth prediction is less than 2 meters , which means it can be used to precisely predict the coring horizon at well site .

这些方法提高了程序代码的并行性和存储器的访问效率，其中重点介绍了算法中比较耗时的半像素插值、绝对误差和（SAD）等关键模块的线性汇编优化。

This paper emphasizes on the linear assembly optimization of some time-consuming key modules such as half pixel interpolate , sum of absolute difference ( SAD ) and so on . Optimized linear assemble language is used to improve real timing and precision .

用速度场建好后完成的两口VSP井实测速度数据检验表明，速度场预测速度最大绝对误差是54m/s，平均为37m/s；

With this velocity field , we test two VSP wells which were completed using velocity field data . The pre dicted max absolute error is54m / s , the average is37m / s ;

标准光栅的测试验证，系统绝对误差小于5m，表明该系统稳定可靠，一致性好。

Test of standard grating demonstrates that the system absolute error is less than 5 microns . This shows that the system is stable and reliable , and has a good consistency .

从相关系数R和回归方程的显著性检验（F）都说明r与Sr密切相关，其中102种阳离子半径数据与具有代表性的离子半径参考值相比平均绝对误差仅0.9pm，相对误差1.1%。

The correlation coefficient R and test of significance ( F ) showed that r was closely related to Sr. The average absolute error of the 102 radii of positive ions was only 0.9 pm with the representative radii of ions , the relative error was 1.1 % .

采用Lorenz方程组研究了非线性系统扰动发展的线性估计和非线性估计，基于扰动的非线性发展计算了系统扰动线性发展的相对误差和绝对误差。

The linear and nonlinear evolutions of initial perturbations in Lorenz system are investigated . According to the nonlinear evolution of initial perturbations , the relative and absolute errors of the perturbation development are computed .

实验结果表明，绝对误差可以控制在0～5mm之间，达到了收获机器人定位精度的要求。

The experiment indicates that the errors can be controlled in the range of 0 - 5 mm , thus meeting the requirement for positioning of robots harvesting fruits .

利用推广的介值定理，得到了求一类方程绝对误差为0.1m（m∈N）的近似解的一种行之有效的方法。

By using the deducted intermediate value theorem , an effective methods of getting approximate solutions to equations of the same class , in which the absolute error is 0.1 m ( m ∈ N ), is obtained .

其中ST段的上升，下降以及起始点和终点的幅值的绝对误差一般小于50μV，相对误差一般小于0.15，重建结果在一定的程度上是可以接受的。

The absolute error of ST rising , lowering part and the absolute errors of amplitudes of ST starting and stopping points are normally all less than 50 μ V.The relative error is less than 0 15 , so the results of the reconstruction are generally acceptable .