Artificial Neural Networks: Approximation and Learning Theory

By Halbert White. 1992

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

Machine Learning Methods That Economists Should Know About (2019, background): “In the third subsection we discuss neural nets, which were the focus on of a small econometrics literature in the 1990s (White [1992], Hornik et al. [1989]), but more recently has become a very prominent literature in ML in various subtle reincarnations.”
Deep Neural Networks for Estimation and Inference: Application to Causal Effects and Other Semiparametric Estimands (2018, background): “In that time, shallow neural networks with smooth activation functions were shown to have many good theoretical properties (White (1992), Anthony and Bartlett (1999), Chen and White (1999)).”
Deep Learning for Mortgage Risk* (2016, background): “5For a broad introduction to deep learning, see White (1992) and Goodfellow, Bengio & Courville (2016).”
The Impact of Machine Learning on Economics (2018, background): “(Also note that White (1992) attempted to popularize neural nets in economics in the early 1990s, but at the time they did not lead to substantial performance improvements and did not become popular in economics.)”
Trends in extreme learning machines: A review (2015, background): “The universal approximation capability of SLFNs has been well studied in the past decades (Poggio & Girosi, 1990; White, 1992).”
Statistical strategies for avoiding false discoveries in metabolomics and related experiments (2007, methodology): “Many of the powerful methods available today, such as neural networks and Evolutionary Algorithms, can usually learn the training set (White, 1992), even if the inputs consist of random numbers.”
Sales forecasting using extreme learning machine with applications in fashion retailing (2008, background): “Recently, artificial neural networks (ANNs) have been applied extensively for sales forecasting [4,13,34,35,44–46] because they have very promising performance in the areas of control, prediction, and pattern recognition [15,21,22,30,33,38,40].”
Cross-sectional forecasts of the equity premium (2006, methodology): “We used the following algorithm, which draws on suggestions in Bishop (1995) (chapter 7), White (1992) (chapter 11) and Masters (1993) (chapter 9).”
Catastrophic Forgetting, Rehearsal and Pseudorehearsal (1995, background): “Common “multilayer perceptron” network architectures such as backpropagation can be conceptualised as function approximators (see White (1992) for an overview).”
EM-based optimization of microwave circuits using artificial neural networks: the state-of-the-art (2003, background): “In principle, they offer an accurate vehicle to model complex phenomena since it has been shown [16] that standard multilayer feed-forward networks can approximate any measurable function to any desired level of accuracy, provided a deterministic relationship between input and target exists.”