Neural net algorithms that learn in polynomial time from examples and queries

By Eric B. Baum. 1991

How researchers see this paper

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

Citations in the corpus

Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data (2016, background): “Active learning is a class of techniques that aims to identify and prioritize points in the student’s training set that have a high potential to contribute to learning (Angluin, 1988; Baum, 1991).”
The Evidence Framework Applied to Classification Networks (1992, background): “This strategy and that of Baum (1991) are both human{designed strategies and it is not clear what objective function if any they optimise, nor is it clear how the strategies could be improved.”
Model Reconstruction from Model Explanations (2018, methodology): “The geometric intuition for our algorithm is similar to the work of Baum for learning two-layer linear threshold networks with membership queries [3].”
Less is More: Active Learning with Support Vector Machines (2000, background): “In special cases, even the computational requirements decrease, and some NP-complete learning problems become polynomial in computation time (Angluin, 1988; Baum & Lang, 1991).”
Loading Deep Networks Is Hard (1994, background): “The polemic of the relevance of the loading problem in the book review by Baum (1991) is also very inspiring.”
Bayesian Methods for Adaptive Models (2011, background): “This strategy and that of Baum [5] are both human–designed strategies and it is not clear what objective function if any they optimise, nor is it clear how the strategies could be improved.”
Information-Based Objective Functions for Active Data Selection (1992, background): “…the neural networks literature on active data selection, also known as`query learning,' has concentrated on slightly diierent problems: The work of Baum (1991) and Hwang et al. (1991) relates to perfectly separable classiication problems only; in both these papers a sensible query{based learning…”
Selective Sampling Using the Query by Committee Algorithm (1997, background): “Baum(Baum,1991), proposed a learning algorithm that uses membership queries to avoid the intractability of learning neural networks with hidden units.”