SynC: A Copula based Framework for Generating Synthetic Data from Aggregated Sources

Li, Zheng; Zhao, Yue; Fu, Jialin

doi:10.1109/icdmw51313.2020.00082

Cited by 22 publications

(15 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Classical approaches for the data generation task include Copulas (Patki et al, 2016;Li et al, 2020b) and Bayesian Networks (Zhang et al, 2017b), where for the latter those based on the approximation proposed by Chow and Liu (1968) are especially popular. In the deep-learning era, Generative Adversarial Networks (GANs) (Goodfellow et al, 2014) have proven highly successful for the generation of images (Radford et al, 2016;Karras et al, 2020).…”

Section: Methodsmentioning

confidence: 99%

Deep Neural Networks and Tabular Data: A Survey

Borisov¹,

Leemann²,

Seßler³

et al. 2021

Preprint

View full text Add to dashboard Cite

Heterogeneous tabular data are the most commonly used form of data and are essential for numerous critical and computationally demanding applications. On homogeneous data sets, deep neural networks have repeatedly shown excellent performance and have therefore been widely adopted. However, their application to modeling tabular data (inference or generation) remains highly challenging. This work provides an overview of state-of-the-art deep learning methods for tabular data. We start by categorizing them into three groups: data transformations, specialized architectures, and regularization models. We then provide a comprehensive overview of the main approaches in each group. A discussion of deep learning approaches for generating tabular data is complemented by strategies for explaining deep models on tabular data. Our primary contribution is to address the main research streams and existing methodologies in this area, while highlighting relevant challenges and open research questions. To the best of our knowledge, this is the first in-depth look at deep learning approaches for tabular data. This work can serve as a valuable starting point and guide for researchers and practitioners interested in deep learning with tabular data.

show abstract

Section: Methodsmentioning

confidence: 99%

Deep Neural Networks and Tabular Data: A Survey

Borisov¹,

Leemann²,

Seßler³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Synthetic data -artificially generated data that mimic the original (observed) data by preserving relationships between variables (Nowok et al, 2016) -may be useful in several areas such as healthcare, finance, data science, and machine learning (Dahmen & Cook, 2019;Kamthe et al, 2021;Nowok et al, 2016;Patki et al, 2016). As such, copula-based data generation models -probabilistic models that allow for the statistical properties of observed data to be modelled in terms of individual behavior and (inter-)dependencies (Joe, 2014) -have shown potential in several applications such as finance, data science, and meteorology (Kamthe et al, 2021;Li et al, 2020;Meyer, Nagler, et al, 2021;Patki et al, 2016). Although copula-based data generation tools have been developed for tabular data -e.g.…”

Section: Summary and Statement Of Needmentioning

confidence: 99%

“…Although copula-based data generation tools have been developed for tabular data -e.g. the Synthetic Data Vault project using Gaussian copulas and generative adversarial networks (Patki et al, 2016;Xu & Veeramachaneni, 2018), or the Synthetic Data Generation via Gaussian Copula (Li et al, 2020) -in computational sciences such as weather and climate, data often consist of large, labelled multidimensional datasets with complex dependencies.…”

Section: Summary and Statement Of Needmentioning

confidence: 99%

Synthia: multidimensional synthetic data generation in Python

Meyer¹,

Nagler²

2021

JOSS

View full text Add to dashboard Cite

“…This is an efficient algorithm that scales well in high dimensional settings. Unlike proximity based models that require pairwise distance calculation [38,26] or learning based models that require training, COPOD incurs low computational overhead.…”

Section: Probabilistic Based Outlier Detection Techniquesmentioning

confidence: 99%

Comparison of Outlier Detection Techniques for Structured Data

Agarwal,

Gupta

2021

Preprint

View full text Add to dashboard Cite

An outlier is an observation or a data point that is far from rest of the data points in a given dataset or we can be said that an outlier is away from the center of mass of observations. Presence of outliers can skew statistical measures and data distributions which can lead to misleading representation of the underlying data and relationships. It is seen that the removal of outliers from the training dataset before modeling can give better predictions. With the advancement of machine learning, the outlier detection models are also advancing at a good pace. The goal of this work is to highlight and compare some of the existing outlier detection techniques for the data scientists to use that information for outlier algorithm selection while building a machine learning model.

show abstract

SynC: A Copula based Framework for Generating Synthetic Data from Aggregated Sources

Cited by 22 publications

References 22 publications

Deep Neural Networks and Tabular Data: A Survey

Deep Neural Networks and Tabular Data: A Survey

Synthia: multidimensional synthetic data generation in Python

Comparison of Outlier Detection Techniques for Structured Data

Contact Info

Product

Resources

About