Adaptive behavior with fixed weights in RNN: an overview

By D. Prokhorov, Lee, Feldkamp, Ivan Yu. Tjukin. 2002

56 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Model Architectures, Signal Processing

How researchers see this paper

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

In Artificial Intelligence

In Model Architectures

In Signal Processing

Citations in the corpus

Deep learning in neural networks: An overview (2014, methodology): “RNNs can also be used for metalearning (Schmidhuber, 1987; Schaul and Schmidhuber, 2010; Prokhorov et al., 2002), because they can in principle learn to run their own weight change algorithm (Schmidhuber, 1993a).”
Deep Neural Network for Structural Prediction and Lane Detection in Traffic Scene (2017, background): “The observed behavior is consistent with what to expect from the well-trained RNN on challenging sequence prediction tasks (see [28], [29]).”
Learning Gradient Descent: Better Generalization and Longer Horizons (2017, methodology): “The update rules of several common optimization algorithms are listed in Table 1.”
Forecasting across time series databases using recurrent neural networks on groups of similar series: A clustering approach (2017, background): “…with a set of fixed weight vectors can still show different predictive behaviour for different time series, i.e., in a trained network, a particular time series is represented by the composition of weight vector and its corresponding internal state (Prokhorov et al., 2002; Smyl and Kuber, 2016).”
Simple and conditioned adaptive behavior from Kalman filter trained recurrent networks (2003, background): “The latter have aspects of orbital attractors but apply to driven rather than autonomous systems; see (Prokhorov et al., 2002) for a related discussion.”
Reservoir computing approaches to recurrent neural network training (2009, unknown): “, the “meta-learning” task in [79] (4).”