The Art of Data Augmentation

By D. V. van Dyk, Xiao-Li Meng. 2001

1,147 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/2ed57da0adc1909c8cba41b077ff780de6c34363

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 Statistics and Probability

Citations in the corpus

Bayesian Methods for Hidden Markov Models (2002, background): “…h is that highly correlated elements of ˆ under p4ˆ d n 1 5 are often nearly independent under p4ˆ d n 1 1 h 5 , so including h in the sampling algorithm provides an expanded parameter space that may accelerate mixing with respect to ˆ if h can be ef(cid:142) ciently drawn (van Dyk and Meng 2001).”
Sufficient burn-in for Gibbs samplers for a hierarchical random effects model (2004, methodology): “As we now explain, this question was addressed by van Dyk and Meng (2001) in a different context.”
Gabor Convolutional Networks (2017, methodology): “Given rich and often redundant convolutional filters, data augmentation is used to achieve local/global transform invariance [26].”
Exact Computation of the Observed Information Matrix for Hidden Markov Models (2002, background): “In this context, the hidden states of the Markov chain are viewed as missing data or as a type of data augmentation (van Dyk and Meng 2001).”
Unbiased Markov chain Monte Carlo methods with couplings (2020, background): “…median of {τj− ξj , j = i} Step 3: compute the final estimator H̄Åk:m as the (sample) average of {HÅk:m.X.i/, Y.i/;ηi/}Ni=1 via efficiency swindles (e.g. Van Dyk and Meng (2001), Craiu and Meng (2001, 2005), Craiu and Lemieux (2007) and Yu and Meng (2011)), among which figures prominently the…”
A comprehensive survey of recent trends in deep learning for digital images augmentation (2021, background): “Data augmentation (Dyk and Meng 2001) is greatly important to overcome the limitation of data samples and particularly image data-sets.”
Bayesian nonparametric hidden semi-Markov models (2012, methodology): “However, we can introduce auxiliary variables to recover conjugacy, following the general data augmentation technique described in (Van Dyk and Meng, 2001).”
A review of ensemble learning and data augmentation models for class imbalanced problems: combination, implementation and evaluation (2023, background): “Data augmentation has the potential to improve the model performance of such models [203].”
Oriented Response Networks (2017, methodology): “Given rich, and often redundant, convolutional ﬁlters, data augmentation can be used to achieve local/global transform invariance [43].”
Convergence rates and asymptotic standard errors for Markov chain Monte Carlo algorithms for Bayesian probit regression (2007, background): “…55, where xi1 and xi2 are covariates indicating the levels of certain antibodies in the ith individual and yi is an indicator for latent membranous lupus nephritis (1 for presence and 0 for absence). van Dyk and Meng (2001) considered the model Pr.Yi =1/=Φ.β0 +β1xi1 +β2xi2/, with a flat prior on β.”
Dreaming More Data: Class-dependent Distributions over Diffeomorphisms for Learned Data Augmentation (2015, background): “This is a key consideration when training deep models, cf. [19].”
Graph-based Isometry Invariant Representation Learning (2017, methodology): “One intuitive way to make the classiﬁcation architectures more robust to isometric transformations is to augment the training set with transformed data (e.g., (Dyk & Meng, 2012)), which however, increases both the training set and training time.”