期刊
INTERNATIONAL JOURNAL OF APPROXIMATE REASONING
卷 140, 期 -, 页码 298-313出版社
ELSEVIER SCIENCE INC
DOI: 10.1016/j.ijar.2021.10.011
关键词
Conditional probabilistic modeling; Probabilistic circuits; Sum-product-networks; Tractable inference; Variational inference
资金
- German Science Foundation (DFG) project Argumentative Machine Learning (CAML) [KE 1686/3-1, SPP 1999]
- Rhine-Main Universities Network for Deep Continuous-Discrete Machine Learning (DeCoDeML)
- DFG [SFB 876]
- European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie Grant [797223]
- Marie Curie Actions (MSCA) [797223] Funding Source: Marie Curie Actions (MSCA)
This paper introduces a new type of conditional SPNs model, which can be used as tractable building blocks of deep probabilistic models and shows competitive performance in various tasks. The research demonstrates that CSPNs can outperform other probabilistic models and have the potential to improve performance in different tasks.
While probabilistic graphical models are a central tool for reasoning under uncertainty in AI, they are in general not as expressive as deep neural models, and inference is notoriously hard and slow. In contrast, deep probabilistic models such as sum-product networks (SPNs) capture joint distributions and ensure tractable inference, but still lack the expressive power of intractable models based on deep neural networks. In this paper, we introduce conditional SPNs (CSPNs)-conditional density estimators for multivariate and potentially hybrid domains-and develop a structure-learning approach that derives both the structure and parameters of CSPNs from data. To harness the expressive power of deep neural networks (DNNs), we also show how to realize CSPNs by conditioning the parameters of vanilla SPNs on the input using DNNs as gate functions. In contrast to SPNs whose highlevel structure can not be explicitly manipulated, CSPNs can naturally be used as tractable building blocks of deep probabilistic models whose modular structure maintains high-level interpretability. In experiments, we demonstrate that CSPNs are competitive with other probabilistic models and yield superior performance on structured prediction, conditional density estimation, auto-regressive image modeling, and multilabel image classification. In particular, we show that employing CSPNs as encoders and decoders within variational autoencoders can help to relax the commonly used mean field assumption and in turn improve performance. (C) 2021 The Authors. Published by Elsevier Inc.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据