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
- knowledge communication
- Knowledge community
- knowledge community
- knowledge competence
- knowledge compilation
- knowledge conflict
- knowledge creation
- knowledge determinism
- knowledge diffusion
- knowledge-directed database
- knowledge discovery
- knowledge discovery from data
- knowledge discovery from data, KDD
- knowledge dissemination
- knowledge-driven decision support system
- knowledge economy
- knowledge editor
- knowledge element
- knowledge employee
- knowledge engineer
- knowledge engineering
- knowledge engineering school
- knowledge evaluation
- knowledge evolution
- knowledge exchange