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
- wide band semiconductor
- wide-band semiconductor
- wide-band signal
- wide beam field curvature
- wide binary
- wide binary galaxy
- wide-body aircraft
- wide class
- wide color gamut liquid crystal display technology
- wide convergence
- wide converter
- wide cross
- wide dynamic
- wide field electromagnetic instrument
- wide field electromagnetic method
- wide-field eyepiece
- wide field imager
- wide-field imaging
- Wide-Field Infrared Explorer
- Wide-field Infrared Survey Explorer
- Wide-Field Infrared Survey Telescope
- wide-field spectrograph
- wide-field telescope
- wide-gap junction
- wide-gap semiconductor