Disentangled VAE Representations for Multi-Aspect and Missing Data

Ainsworth, Samuel; Foti, Nicholas J.; Fox, Emily B.

doi:10.48550/arxiv.1806.09060

Cited by 4 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another alternative for imputing missing data is to rely on unsupervised methods. These methods are based on variational autoencoders (VAEs) [60], [61] or generative adversarial networks (GANs) [62], [63]. In [64], a generic framework for missing data imputation on time series is developed by combining concepts from VAEs [65], Cauchy kernels [66], Gaussian Processes [67], structured variational distributions with efficient inference [68] and a particular Evidence Lower Bound (ELBO) for missing data [61].…”

Section: Reconstruction Methods a Scope And Backgroundmentioning

confidence: 99%

Infinite Limits of Convolutional Neural Network for Urban Electromagnetic Field Exposure Reconstruction

Mallik,

Allaert,

Egea-Lopez

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Electromagnetic field exposure (EMF) has grown to be a critical concern as a consequence of the ongoing installation of fifth-generation cellular networks (5G). The lack of measurements makes it difficult to accurately assess the EMF in a specific urban area, as Spectrum cartography (SC) relies on a set of measurements recorded by spatially distributed sensors for the generation of exposure maps. However, when the spatial sampling rate is limited, significant estimation errors. To overcome this issue, the exposure map estimation is addressed as a missing data imputation task. We compute a convolutional neural tangent kernel (CNTK) for an infinitely wide convolutional neural network whose training dynamics can be completely described by a closed-form formula. This CNTK is employed to impute the target matrix and estimate EMF exposure from a few sensor-measured values located in an urban environment. Experimental results show that the kernel, even when only sparse sensor data are available, can produce accurate estimates. It is a promising solution for exposure map reconstruction that does not require large training sets. The proposed method is compared with other machine learning approaches and Gaussian Process regression.INDEX TERMS 5G EMF exposure, kernel regression, neural tangent kernel, infinite width convolutional neural network, semi-supervised learning.

show abstract

Section: Reconstruction Methods a Scope And Backgroundmentioning

confidence: 99%

Infinite Limits of Convolutional Neural Network for Urban Electromagnetic Field Exposure Reconstruction

Mallik,

Allaert,

Egea-Lopez

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The researchers in [62]- [64] propose the use of non-linear dimensionality reduction to handle missing values in time series using variational autoencoders (VAE). GP-VAE proposed by [36] is especially focused on MTS.…”

Section: ) Vae-gan Modelsmentioning

confidence: 99%

Multivariate Time Series Imputation: A Survey on available Methods with a Focus on hybrid GANs

Richter¹,

Ijaradar²,

Wetzker³

et al. 2022

Preprint

View full text Add to dashboard Cite

<p>Multivariate time series (MTS) are captured in a great variety of real-world applications. However, analysing and modelling the data for classification and forecasting purposes can become very challenging if values are missing in the data set. The need for imputation methods, to fill the gaps in MTS, is well known. Thus, a great variaty of algorithms for solving this task has been proposed in the literature. However, research community is constantly working on the development of advanced algorithms, that fulfill the special requirements of multidimensional temporal data, since most of the existing imputation methods treat MTS as ordinary structured data and fail to model the temporal relationships within and between sequences of observations. The main emphasis of MTS imputation research is currently put on deep learning (DL) models, especially models making use of generative adversarial networks (GANs). In our survey, we present a general categorization of imputation algorithms and introduce groups of hybrid GAN-models used for the MTS imputation task, which we investigate and discuss in detail. A quantitative comparison of the hybrid GANs’ performance regarding MTS imputation is presented based on our findings in the literature. </p>

show abstract

“…Variational auto-encoders (VAEs) have been proposed for several problems within this definition of unsupervised reconstruction [10,5,7]. These methods lead to good single estimates of the underlying targets.…”

Section: Vaes and The Posterior Collapse Problemmentioning

confidence: 99%

“…Finally, Bayesian LVM methods have been used on other unsupervised tasks that can be cast as special cases of data recovery problems. Amongst these, we find Multi-view generation [30,7], where the target clean data includes all views for each samples, but the observed data only presents subsets. Blind source separation can also be cast as a recovery problem and has been approached with GANs and VAEs [31,32].…”

Section: Unsupervised Bayesian Reconstructionmentioning

confidence: 99%

“…Several variational auto-encoder (VAE) models have been proposed for applications that can be considered special cases of this problem [5,6,7] and, in principle, they are capable of performing Bayesian reconstruction. However, we show that surrogating variational inference (VI) in a latent space with VAEs results in collapsed distributions that do not explore the different possibilities of clean samples, but only return single estimates.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Tomographic Auto-Encoder: Unsupervised Bayesian Recovery of Corrupted Data

Tonolini,

Moreno,

Damianou

et al. 2020

Preprint

View full text Add to dashboard Cite

We propose a new probabilistic method for unsupervised recovery of corrupted data. Given a large ensemble of degraded samples, our method recovers accurate posteriors of clean values, allowing the exploration of the manifold of possible reconstructed data and hence characterising the underlying uncertainty. In this setting, direct application of classical variational methods often gives rise to collapsed densities that do not adequately explore the solution space. Instead, we derive our novel reduced entropy condition approximate inference method that results in rich posteriors. We test our model in a data recovery task under the common setting of missing values and noise, demonstrating superior performance to existing variational methods for imputation and de-noising with different real data sets. We further show higher classification accuracy after imputation, proving the advantage of propagating uncertainty to downstream tasks with our model.

show abstract

Disentangled VAE Representations for Multi-Aspect and Missing Data

Cited by 4 publications

References 9 publications

Infinite Limits of Convolutional Neural Network for Urban Electromagnetic Field Exposure Reconstruction

Infinite Limits of Convolutional Neural Network for Urban Electromagnetic Field Exposure Reconstruction

Multivariate Time Series Imputation: A Survey on available Methods with a Focus on hybrid GANs

Tomographic Auto-Encoder: Unsupervised Bayesian Recovery of Corrupted Data

Contact Info

Product

Resources

About