1
哪种深度学习模型可以对不互斥的类别进行分类
示例:我的职位描述中有一句话:“英国Java高级工程师”。 我想使用深度学习模型将其预测为2类:English 和IT jobs。如果我使用传统的分类模型,则只能预测softmax最后一层具有功能的标签。因此,我可以使用2个模型神经网络来预测两个类别的“是” /“否”,但是如果我们有更多类别,那就太贵了。那么,我们是否有任何深度学习或机器学习模型可以同时预测2个或更多类别? “编辑”:使用传统方法使用3个标签,它将由[1,0,0]编码,但在我的情况下,它将由[1,1,0]或[1,1,1]编码 示例:如果我们有3个标签,并且所有这些标签都适合一个句子。因此,如果softmax函数的输出为[0.45,0.35,0.2],我们应该将其分类为3个标签或2个标签,或者可以是一个?我们这样做的主要问题是:分类为1个,2个或3个标签的最佳阈值是多少?
9
machine-learning
deep-learning
natural-language
tensorflow
sampling
distance
non-independent
application
regression
machine-learning
logistic
mixed-model
control-group
crossover
r
multivariate-analysis
ecology
procrustes-analysis
vegan
regression
hypothesis-testing
interpretation
chi-squared
bootstrap
r
bioinformatics
bayesian
exponential
beta-distribution
bernoulli-distribution
conjugate-prior
distributions
bayesian
prior
beta-distribution
covariance
naive-bayes
smoothing
laplace-smoothing
distributions
data-visualization
regression
probit
penalized
estimation
unbiased-estimator
fisher-information
unbalanced-classes
bayesian
model-selection
aic
multiple-regression
cross-validation
regression-coefficients
nonlinear-regression
standardization
naive-bayes
trend
machine-learning
clustering
unsupervised-learning
wilcoxon-mann-whitney
z-score
econometrics
generalized-moments
method-of-moments
machine-learning
conv-neural-network
image-processing
ocr
machine-learning
neural-networks
conv-neural-network
tensorflow
r
logistic
scoring-rules
probability
self-study
pdf
cdf
classification
svm
resampling
forecasting
rms
volatility-forecasting
diebold-mariano
neural-networks
prediction-interval
uncertainty