该方法依据极大似然原理将来自不同母体(均值相同、方差不同)的随机样本有效融合，得到新的母体均值估计量。
According to maximum likelihood theory, it fuses random samples coming from different matrix (same mean different variance) in an effective way, and gains a nwe estimator of matrix mean.
对不能识别的样本，采用模糊推理技术，把传统的直观特征识别结果和多层BP网络结果在特征级上融合，提高系统的性能。
As to the samples that have not been able to be recognized we adopt fuzzy logic to fusion obvious feature and MBPNN's feature, and increase the performance of the system.
方法：样本包括经历孤立的后路脊椎融合和器械固定的青少年特发性侧凸病人。
Methods. The sample consisted of patients with AIS who underwent isolated posterior spinal fusion and instrumentation.
在检索特定领域信息时，通过相关样本集融合，提取出关键词集，通过调节样本集实现关键词集的柔性控制，以调控搜索空间与结果取向。
When searching special area information, it fuses sample documents, extracts common keywords, and adjusts the sample documents to control the search space and result sum.
训练样本与测试样本分别朝融合特征空间投影，从而得到识别特征。
After training samples and test samples are respectively projected towards the fusion feature space, recognition features are accordingly gained.
基于该融合公式，证明了在单传感器系统中，当样本观测值一定时，分组数据融合的估计效果优于单组算术平均的估计效果；
In a single sensor system when the sample amount is constant, the estimation result of data fusion to subdivided into groups is superior to arithmetic average of single group.
将融合后的最终结果输入到高阶BP神经网络中，通过目标向量样本的训练，输出相应的目标类型。
Then the final results into the high order BP network is input. Through the training of target vector table stylebook, corresponding type of goal can be output.
为了解决在没有已知标签样本的情况下数据流组合分类决策问题，提出一种基于约束学习的数据流组合分类器的融合策略。
To resolve combining classifiers decisions among ensemble classification over data streams without labeled examples, a transductive constraint-based learning strategy was proposed.
为了解决在没有已知标签样本的情况下数据流组合分类决策问题，提出一种基于约束学习的数据流组合分类器的融合策略。
To resolve combining classifiers decisions among ensemble classification over data streams without labeled examples, a transductive constraint-based learning strategy was proposed.
应用推荐