A Framework for the Cooperation of Learning Algorithms

By L. Eon Bottou, Patrick Gallinari. 1990

Subfields: Reinforcement Learning, Meta & Continual Learning

How researchers see this paper

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

In Reinforcement Learning

In Meta & Continual Learning

Citations in the corpus

Natural Language Processing (Almost) from Scratch (2011, background): “We report for each task the window size (or convolution size), word feature dimension, capital feature dimension, number of hidden units and learning rate. compute derivatives with respect to its inputs and with respect to its trainable parameters, as proposed by Bottou and Gallinari (1991).”
DeepArchitect: Automatically Designing and Training Deep Architectures (2017, methodology): “This is akin to the concept of a module [6] used in deep learning frameworks such as Torch [8]: by implementing the module interface, the internal implementation becomes irrelevant.”
Object Files and Schemata: Factorizing Declarative and Procedural Knowledge in Dynamical Systems (2020, background): “ding to ﬁxed dynamics and each RIM module is specialized for a particular computation. RIM modules are thus not interchangeable. Neural Module Networks. A modular module network [Jacobs et al., 1991, Bottou and Gallinari, 1991, Ronco et al., 1997, Reed and De Freitas, 2015, Andreas et al., 2016, Ros”
Recurrent Independent Mechanisms (2019, background): “…neural modules, where each module is meant to perform a distinct function, and hence can be seen as a combination of experts (Jacobs et al., 1991; Bottou & Gallinari, 1991; Ronco et al., 1997; Reed & De Freitas, 2015; Andreas et al., 2016; Parascandolo et al., 2018; Rosenbaum et al., 2017;…”
Coordination Among Neural Modules Through a Shared Global Workspace (2021, background): “While modular architectures have been the subject of a number of research directions, (Jacobs et al., 1991; Bottou and Gallinari, 1991; Ronco et al., 1997; Reed and De Freitas, 2015; Andreas et al., 2016; Rosenbaum et al., 2017; Fernando et al., 2017; Shazeer et al., 2017; Rosenbaum et al., 2019;…”
Global training of document processing systems using graph transformer networks (1997, methodology): “The concept of building trainable systems by assembling modules and propagating gradients through them has been proposed for quite some time [3], and used extensively in practical systems [2].”
Neural Production Systems (2021, background): “…be composed of several modules, each meant to perform a distinct function, and hence can be seen as a combination of experts (Jacobs et al., 1991; Bottou & Gallinari, 1991; Ronco et al., 1997; Reed & De Freitas, 2015; Andreas et al., 2016; Rosenbaum et al., 2017; Fernando et al., 2017; Shazeer…”
Deep Online Convex Optimization by Putting Forecaster to Sleep (2015, background): “The key technical innovation is the notion of a path-sum game constructed over a directed acyclic graph with weighted edges.”
Gradient-based learning applied to document recognition (1998, background): “Learning these weight con(cid:12)gurations requires a very large number of training instances to cover the space of possible variations.”
Fast and Slow Learning of Recurrent Independent Mechanisms (2021, background): “…of models, to make them resemble the kind of rules they are supposed to learn (Andreas et al., 2016; Johnson et al., 2017; Jacobs et al., 1991; Bottou & Gallinari, 1991; Ronco et al., 1997; Reed & De Freitas, 2015; Andreas et al., 2016; Rosenbaum et al., 2017; Fernando et al., 2017; Shazeer…”
Modular Neural Networks and Self-Decomposition (1997, background): “As (Mure et al. 1992) (Jacobs et al. 1991a) (Bottou and Galliinari 1991)highlight, modularity is a way to embed in NN a priori knowledge that is knownto be important to improve NN learning (LeCun 1987).”
Object Recognition with Gradient-Based Learning (1999, methodology): “A completely rigorous derivation of the gradient propagation procedure in the general case can be done using Lagrange functions [14, 15, 2 ]. Object Recognition with Gradient-Based Learning 321”