paper-with-me

홈 › Papers

Information theoretic learning of robust deep representations

2019-05-30 · Nicolas Pinchaud

We propose a novel objective function for learning robust deep representations of data based on information theory. Data is projected into a feature-vector space such that the mutual information of all subsets of features relative to the supervising signal is maximized. This objective function gives rise to robust representations by conserving available information relative to supervision in the face of noisy or unavailable features. Although the objective function is not directly tractable, we are able to derive a surrogate objective function. Minimizing this surrogate loss encourages features to be non-redundant and conditionally independent relative to the supervising signal. To evaluate the quality of obtained solutions, we have performed a set of preliminary experiments that show promising results.

📄 PDF Abstract BibTeX arXiv:1905.12874

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

INFORMATION MAXIMIZATION AUTO-ENCODING

2019-05-01 · ICLR 2019 5 · Dejiao Zhang, Tianchen Zhao, Laura Balzano

We propose the Information Maximization Autoencoder (IMAE), an information theoretic approach to simultaneously learn continuous and discrete representations in an unsupervised setting. Unlike the Variational Autoencoder…

DecoderDisentanglementInformativenessvalid

Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach

2021-05-06 · ACL 2021 5 · Yifan Hou, Mrinmaya Sachan

NLP has a rich history of representing our prior understanding of language in the form of graphs. Recent work on analyzing contextualized text representations has focused on hand-designed probe models to understand how a…

Towards Comprehensive Information-theoretic Multi-view Learning

2025-09-02 · Long Shi, Yunshan Ye, Wenjie Wang, Tao Lei 외 arxiv

Information theory has inspired numerous advancements in multi-view learning. Most multi-view methods incorporating information-theoretic principles rely an assumption called multi-view redundancy which states that commo…

Representation Learning

Which Mutual-Information Representation Learning Objectives are Sufficient for Control?

2021-06-14 · NeurIPS 2021 12 · Kate Rakelly, Abhishek Gupta, Carlos Florensa, Sergey Levine

Mutual information maximization provides an appealing formalism for learning representations of data. In the context of reinforcement learning (RL), such representations can accelerate learning by discarding irrelevant a…

Reinforcement Learning (RL)Representation Learning

Information Theory-Guided Heuristic Progressive Multi-View Coding

2021-09-06 · Jiangmeng Li, Wenwen Qiang, Hang Gao, Bing Su 외

Multi-view representation learning captures comprehensive information from multiple views of a shared context. Recent works intuitively apply contrastive learning (CL) to learn representations, regarded as a pairwise man…

Contrastive LearningMULTI-VIEW LEARNINGRepresentation Learning