Mutual information maximization: models of cortical self-organization.

By Suzanna Becker. 1996

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Theory

Citations in the corpus

Learning deep representations by mutual information estimation and maximization (2018, background): “Leveraging known structure in the input in MI maximization-based objectives is nothing new (Becker, 1992, 1996; Wiskott & Sejnowski, 2002), and some very recent works also follow this intuition.”
Deep learning from temporal coherence in video (2009, background): “The work of Becker (1996a; 1999), which is probably the most related work to ours, explores the use of temporal context using a fully connect neural network algorithm which introduces extra neurons, called contextual gating units, and a Hebbian update rule for clustering based on context (…”
Slow Feature Analysis: Unsupervised Learning of Invariances (2002, unknown): “For an overview, see (Becker 1996).”
Multi-view Regression Via Canonical Correlation Analysis (2007, background): “The Imax framework of Becker and Hinton [1992], Becker [1996] attempts to maximize information between two views x(1) and x(2), for which CCA is a special case (in a continuous version).”
Information Bottleneck for Gaussian Variables (2003, methodology): “Keywords: information bottleneck, Gaussian processes, dimensionality reduction, canonical correlation analysis”
Deep multi-view learning methods: A review (2021, background): “[99] also investigate to extract high order features from multiple views by maximally preserving the mutual information between multiple 310 neural networks.”
A Survey of Multi-View Representation Learning (2016, unknown): “Further, Becker [104] explores the idea of maximizing the mutual information between the outputs of different network modules to extract higher order features from coherence inputs.”