-
Notifications
You must be signed in to change notification settings - Fork 2.2k
Chinese translation
Yang Zhou edited this page Feb 5, 2021
·
171 revisions
File name | Translator | Start date | Publish date | Reviewers |
---|---|---|---|---|
README.md |
jiqihumanR | [x] | Mingyang | |
index.md |
Mingyang | [x] | ||
01.md |
Mingyang | [x] | jiqihumanR | |
01-1.md |
lejinghu, Mingyang | [x] | zhounan | |
01-2.md |
jiqihumanR | [x] | lejinghu | |
01-3.md |
titusjgr | [x] | Lin | |
lecture01.sbv |
||||
practicum01.sbv |
||||
02.md |
Matt | [x] | Yusu | |
02-1.md |
eric | [x] | Mingyang | |
02-2.md |
titusjgr | [x] | Shengkun Tang | |
02-3.md |
titusjgr | [x] | Shengkun Tang | |
lecture02.sbv |
||||
practicum02.sbv |
||||
03.md |
jiqihumanR | [x] | Mingyang ⌛ | |
03-1.md |
jiqihumanR | [x] | Mingyang ⌛ | |
03-2.md |
Mingyang | [x] | ||
03-3.md |
Matt | [x] | Mingyang ⌛ | |
lecture03.sbv |
||||
practicum03.sbv |
||||
04.md |
jiqihumanR | [x] | ||
04-1.md |
jiqihumanR | [x] | ||
practicum04.sbv |
||||
05.md |
Shengkun Tang | [x] | ||
05-1.md |
Shengkun Tang | [x] | ||
05-2.md |
Harry Yang | [x] | ||
05-3.md |
titusjgr | 2020-04-04 | 2020-04-13 | |
lecture05.sbv |
||||
practicum05.sbv |
||||
06.md |
jiqihumanR | 2020-04-17 | [x] | |
06-1.md |
jiqihumanR | 2020-04-17 | ||
06-2.md |
Jonathan Sum | 2020-06-10 | 2020-06-20 | |
06-3.md |
Elizabeth Zhao | 2020-06-11 | 2020-06-18 | |
07.md |
Jonathan Sum | 2020-06-10 | 2020-06-11 | |
07-1.md |
jiqihumanR | 2020-04-17 | ||
07-2.md |
Jonathan Sum | 2020-06-10 | 2020-06-19 | |
07-3.md |
Jonathan Sum | 2020-06-10 | 2020-06-20 | |
08.md |
Jonathan Sum | 2020-06-10 | 2020-06-15 | |
08-1.md |
Jonathan Sum | 2020-06-10 | 2020-06-15 | |
08-2.md |
Jonathan Sum | 2020-06-10 | 2020-06-15 | |
08-3.md |
Elizabeth Zhao | 2020-06-20 | 2020-06-23 | |
09.md |
Teresa Huang | |||
09-1.md |
Teresa Huang | |||
09-2.md |
Zeping Zhan | |||
09-3.md |
Jonathan Sum | 2020-06-10 | 2020-06-15 | |
10-1.md |
Jonathan Sum | 2020-06-20 | 2020-06-27 | |
10.md |
Elizabeth Zhao | 2020-06-29 | 2020-06-29 | |
10-1.md |
Jonathan Sum | 2020-06-20 | 2020-06-27 | |
10-2.md |
Elizabeth Zhao | 2020-07-03 | ||
10-3.md |
Elizabeth Zhao | 2020-06-25 | 2020-07-02 | |
11.md |
titusjgr | 2020-09-03 | 2020-09-04 | |
11-1.md |
Jonathan Sum | 2020-07-01 | 2020-07-06 | |
11-2.md |
Che Wang | 2020-09-18 | ||
11-3.md |
Jonathan Sum | 2020-07-05 | 2020-07-10 | |
12.md |
Jonathan Sum | 2020-09-02 | 2020-09-04 | |
12-1.md |
eric | 2020-06-27 | ||
12-2.md |
Shengkun Tang | 2020-07-16 | ||
12-3.md |
Shengkun Tang | 2020-08-30 | ||
13.md |
Shengkun Tang | 2020-09-03 | ||
13-1.md |
Yusu | |||
13-2.md |
Jonathan Sum | 2020-09-04 | 2020-09-16 | |
13-3.md |
Jonathan Sum | 2020-09-16 | 2020-09-17 | |
14.md |
Jonathan Sum | 2020-09-03 | 2020-09-04 | |
14-1.md |
titusjgr | 2020-09-18 | ||
14-2.md |
Yang Zhou | 2021-01-31 | 2021-01-31 | |
14-3.md |
Yang Zhou | 2021-02-03 | 2021-02-03 | |
15.md |
Yang Zhou | 2021-02-04 | 2021-02-04 | |
15-1.md |
Yang Zhou | 2021-02-04 | 2021-02-04 | |
15-2.md |
Yang Zhou | 2021-02-04 | 2021-02-05 |
Term | Targeted Chinese |
---|---|
activation | 激活值 |
activation function | 激活函数 |
activation-selective neuron | 选择性激活神经元 |
actor critic | 演员-评判家 |
affine transformation | 仿射变换 |
argument(programming) | 实参 |
(artificial) neural network | (人工)神经网络 |
attention | 注意力机制 |
autonomous vehicles | 自动驾驶(技术) |
back propagation | 反向传播(算法) |
batch | 批次 |
batch normalization | 批量归一化 |
batch size | 批次大小 |
bias | 偏置 |
chain rule | 链式法则 |
combinational logic | 组合电路 |
contrast normalization | 对比度归一化 |
contrastive learning | 对比学习 |
convergence | 收敛 |
convolution | 卷积 |
cost function | 代价函数 |
cybernetic | 控制论 |
deep-learning | 深度学习 |
digital electronics | 数字电路 |
discriminative model | 判别模型 |
distortion | 畸变 |
divergence | 发散 |
embedding | 嵌入 |
Energy-Based Models | 能量基础模型 |
Energy-Based Factor Graph | 能量基础因素图形 |
ensemble | 集成 |
epoch | 周期 |
explicit | 显性的 |
flip-flop | 触发器 |
forward algorithm | 前向式算法 |
forward pass | 前向传播 |
fully connected layer | 全连接层 |
generative model | 生成模型 |
gradient descent | 梯度下降(算法) |
Graph Transformer Networks, GTNs | 图形式变压器网路 |
hierarchical representation | 层次结构 |
hypernetwork | 超网络 |
image classification | 图像分类 |
image segmentation | 图像分割 |
implicit | 隐性的 |
inductive bias | 归纳偏置 |
inference | 推断 |
Jacobian matrix | 雅可比矩阵 |
lane tracking | 车道跟踪 |
latent space | 隐空间 |
latent variable | 隐变量 |
learning rate | 学习率 |
logistic regression | 逻辑回归 |
loss function | 损失函数 |
Nash equilibrium | 纳什均衡 |
nearest neighbor | 近邻算法 |
object detection | 目标检测 |
parameters(programming) | 形参 |
pattern recognition | 模式识别 |
penalty term | 惩罚因子 |
pooling | 池化(下采样) |
receptive field | 感受野 |
reflection | 反射 |
regularization | 正则化 |
rotation | 旋转 |
scalar | 标量 |
scaling | 缩放 |
self-supervised learning | 自监督学习 |
semantic segmentation | 语义分割 |
sequential data | 序列数据 |
sequential logic | 时序逻辑电路 |
sequential model | 序列模型 |
shearing | 错切 |
stochastic gradient descent | 随机梯度下降法 |
supervised learning | 监督学习 |
TDNN | 时延神经网络 |
tensor | 张量 |
translation | 平移 |
trajectory | 轨迹 |
unsupervised learning | 无监督学习 |
vanilla | 基础型 |
visual cortex | 视觉皮层 |
weight | 权重 |
weight decay | 权重衰减 |
weighted sum | 加权和 |
Term | Targeted Chinese |
---|---|
Kunihiko Fukishima | 福岛邦彦 |
Term | Explanation |
---|---|
ClusterFit | ClusterFit |
CNN | Convolutional Neural Network 卷积网络 |
dropout | dropout |
GAN | generative adversarial networks 生成对抗网络 |
GPU | GPU |
Jupyter Notebook | An open-source live document |
one-hot | one-hot |
pretext task | Did not find a formal translation |
PyTorch | PyTorch |
ReLU | Rectified linear unit 修正线性单元 |
Transformer | Transformer |
VAE | Variational Auto-Encoder 变分自编码器 |
Vanilla NN | Vanilla Neural Network ("vanilla" means "plain") |