理学 随机梯度下降 释 stochastic gradient descent; SGD 随机梯度下降 依据损失函数的梯度，在一次迭代中对每一个训练实例进行预测，并重复迭代该过程到一定的次数，从而最小化一个函数的过程。在机器学习中，可以利用该方法来最小化训练模型中的误差，即每次迭代时完成一次评估和更新。
人工智能 神经网络 释文 stochastic gradient descent 随机梯度下降法
- water metabolism
- water meters
- water migration
- water milfoil family
- water mist fire extinguishing system
- water mite
- water model
- water models
- watermoss genus
- watermosss
- water mould
- water neutron shield
- water nutrition
- water of constitution
- water of crystallization
- Water of Rivers And Lakes
- water oil dough
- water on Mars
- water organic pollutants analysis
- Water-out layer
- water park
- water pathogen attacking heart
- water permeable pavement recovery machine
- water permeation consistent of cocoon
- water phantom