Artificial Intelligence
贝叶斯神经网络
Bayesian neural networksA neural network that incorporates Bayesian inference to provide probabilistic predictions rather than deterministic outputs. It quantifies uncertainty in model parameters and predictions. / 一种结合贝叶斯推断的神经网络，提供概率性预测而非确定性输出。它量化了模型参数和预测中的不确定性。
- Knowledge type
- 数学理论
- First proposed or discovered
- Proposer
- Turkey Farm member
- Creator
- 农场主
- Last updated
- 2026-10-29 18:05:59
Sub-knowledge points
Variational inferenceMarkov Chain Monte Carlo (MCMC)Uncertainty quantification
Higher-order knowledge points
Probabilistic graphical modelsDeep learning theory
Main parameters and performance indicators
Posterior distribution: The updated probability distribution of model parameters after observing data.Prior distribution: The initial probability distribution of model parameters before observing data.Likelihood function: The probability of observing the data given the model parameters.
Open problems
Computational complexity in training deep Bayesian networks.Scalability to very large datasets.
Related literature
ranked by community starsPhysics-constrained Bayesian neural networks for probabilistic fatigue life prediction of glass fabric polymer compositesOpen source
Related products and links
No item recorded yet
Creation and modification trail
1 recordsRecord title: Bayesian neural networks.
Node update feedback
review required