Sub-sampling

Sub-sampling

The Analysis of Variance for Designed Experiments

Published in William M. Mendenhall, Terry L. Sincich, Statistics for Engineering and the Sciences, 2016

William M. Mendenhall, Terry L. Sincich

One way to separate and to estimate the magnitudes of these two sources of variation is to perform the sampling in two stages. First, we randomly select n1 specimens from the batch. Then we measure the potency of each specimen n2 times. Because n2 second-stage sampling units are obtained from each first-stage or primary unit (see Figure 14.26), the sampling procedure is called a nested sampling design. It is also referred to as subsampling—that is, sampling within a sample.

The image-based analysis and classification of urine sediments using a LeNet-5 neural network

View Article

Journal Information

Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2020

Taihao Li, Di Jin, Cuifen Du, Xiumei Cao, Haige Chen, Jianshe Yan, Na Chen, Zhenyi Chen, Zhenzhen Feng, Shupeng Liu

A sub-sampling layer performs sub-sampling on its input, based on the principle of local correlation of images. Sub-sampling can reduce the amount of data to process while retaining useful information. If the number of feature maps in the input is , then the number of feature maps in the output is also , but each feature map has a reduced size, e.g., after one sub-sampling operation, the output feature map can be of a quarter of the input feature map. Sub-sampling can be described by Equation (2)

Explore chapters and articles related to this topic

The Analysis of Variance for Designed Experiments

The image-based analysis and classification of urine sediments using a LeNet-5 neural network