Open Access
表1
训练伪代码
输入: 训练集Dbase={(xi, yi)}i=1n, yi∈{1, 2, …, N}, 其中xi表示信号样本, yi表示样本标签 |
过程: |
step 1: 从含有N个类别的Dbase中随机选取C类样本组成样本集DV用于构建DTS和DTQ |
step 2: For k in {1, …, NC} do |
step 3: 从DV中的第k类样本中随机抽取K个样本组成DTSk |
step 4: 从DV-DTSk中的第k类样本中随机抽取NQ个样本组成DTQk |
Step 5: DTS←DTSk |
Step 6: DTQ←DTQk |
Step 7: ![]() |
Step 8: End for |
Step 9: J(φ)=0 |
Step 10: For k in {1, …, NC} do |
Step 11: For (x, y) in DTQk do |
Step 12: ![]() |
Step 13: End for |
Step 14: End for |
输出: 训练损失J(φ) |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.