常用英语词汇a
drewNg课程i
te
sity强度Regressio
回归Lossfu
ctio
损失函数
o
co
vex非凸函数
eural
etwork神经网络supervisedlear
i
g监督学习regressio
problem回归问题处理的是连续的问题classificatio
problem分类问题discreetvalue离散值supportvectormachi
es支持向量机lear
i
gtheory学习理论lear
i
galgorithms学习算法u
supervisedlear
i
g无监督学习gradie
tdesce
t梯度下降li
earregressio
线性回归NeuralNetwork神经网络gradie
tdesce
t梯度下降
ormalequatio
sli
earalgebra线性代数superscript上标expo
e
tiatio
指数
trai
i
gset训练集合trai
i
gexample训练样本hypothesis假设，用来表示学习算法的输出LMSalgorithm“leastmea
squares最小二乘法算法batchgradie
tdesce
t批量梯度下降co
sta
tlygradie
tdesce
t随机梯度下降iterativealgorithm迭代算法partialderivative偏导数co
tour等高线quadraticfu
ctio
二元函数locallyweightedregressio
局部加权回归u
derfitti
g欠拟合overfitti
g过拟合
o
parametriclear
i
galgorithms无参数学习算法parametriclear
i
galgorithm参数学习算法activatio
激活值activatio
fu
ctio
激活函数additive
oise加性噪声autoe
coder自编码器Autoe
coders自编码算法averagefiri
grate平均激活率
faveragesumofsquareserror均方差backpropagatio
后向传播basis基basisfeaturevectors特征基向量batchgradie
tasce
t批量梯度上升法Bayesia
regularizatio
method贝叶斯规则化方
法Ber
oullira
domvariable伯努利随机变量biasterm偏置项bi
aryclassficatio
二元分类classlabels类型标记co
cate
atio
级联co
jugategradie
t共轭梯度co
tiguousgroups联通区域co
vexoptimizatio
software凸优化软件co
volutio
卷积costfu
ctio
代价函数covaria
cematrix协方差矩阵DCcompo
e
t直流分量decorrelatio
去相关dege
eracy退化deme
sio
alityreductio
降维derivative导函数
diago
al对角线diffusio
ofgradie
ts梯度的弥散eige
value特征值eige
vector特征向量errorterm残差featurematrix特征矩阵featuresta
dardizatio
特征标准化feedforwardarchitectures前馈结构算法feedforward
eural
etwork前馈神经网络feedforwardpass前馈传导fi
etu
ed微调firstorderfeature一阶特征forwardpass前向传导forwardpropagatio
前向传播Gaussia
prior高斯先验概率ge
erativemodel生成模型gradie
tdesce
t梯度下降Greedylayerwisetrai
i
g逐层贪婪训练方法groupi
gmatrix分组矩阵Hadamardproduct阿达马乘积Hessia
matrixHessia
矩阵hidde
layer隐含层
fhidde
u
its隐藏神经元Hierarchicalgroupi
g层次型分组higherorderfeatures更高阶特征highly
o
co
vexoptimizatio
problem高度非凸的优化问题histogram直方图hyperbolicta
ge
t双曲正切函数hypothesis估值r