指出子空间信息准则是模型选择的一种新准则,它在一些假设条件下,给出推广误差的一种无偏估计。
Subspace information criterion is a new criterion for model selection, it gives an unbiased estimate for the generalization error under some assumptions.
理论上给出了算法的流大小的无偏估计,相对误差的上界以及时间和空间复杂度。
Unbiased estimation of flow size, the upper bound of relative error and the time and space complexity are deduced theoretically.
实验结果表明,该算法平均误差率较小,无偏性较好,在数据缺失较为严重的情况下也能完成填补。
Experimental results show that it has a small average error rate and good unbiasedness, it can still complete data under severe missing circumstance.
讨论多阶段抽样比率估计及其样本量选择的问题,给出了比率估计抽样均方误差的近似公式及其渐近无偏估计的公式。
We give the approximation formulas of the variance of the sample ratio estimator for a population mean and its asymptotically non-biased estimator in multi-stage sampling.
讨论多阶段抽样比率估计及其样本量选择的问题,给出了比率估计抽样均方误差的近似公式及其渐近无偏估计的公式。
We give the approximation formulas of the variance of the sample ratio estimator for a population mean and its asymptotically non-biased estimator in multi-stage sampling.
应用推荐