We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: IGANI: Iterative Generative Adversarial Networks for Imputation with Application to Traffic Data

Abstract: Increasing use of sensor data in intelligent transportation systems calls for accurate imputation algorithms that can enable reliable traffic management in the occasional absence of data. As one of the effective imputation approaches, generative adversarial networks (GANs) are implicit generative models that can be used for data imputation, which is formulated as an unsupervised learning problem. This work introduces a novel iterative GAN architecture, called Iterative Generative Adversarial Networks for Imputation (IGANI), for data imputation. IGANI imputes data in two steps and maintains the invertibility of the generative imputer, which will be shown to be a sufficient condition for the convergence of the proposed GAN-based imputation. The performance of our proposed method is evaluated on (1) the imputation of traffic speed data collected in the city of Guangzhou in China, and the training of short-term traffic prediction models using imputed data, and (2) the imputation of multi-variable traffic data of highways in Portland-Vancouver metropolitan region which includes volume, occupancy, and speed with different missing rates for each of them. It is shown that our proposed algorithm mostly produces more accurate results compared to those of previous GAN-based imputation architectures.
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO)
Cite as: arXiv:2008.04847 [stat.ML]
  (or arXiv:2008.04847v3 [stat.ML] for this version)

Submission history

From: Hadi Meidani [view email]
[v1] Tue, 11 Aug 2020 16:46:02 GMT (425kb,D)
[v2] Wed, 20 Jan 2021 19:47:56 GMT (2022kb,D)
[v3] Mon, 21 Jun 2021 16:37:58 GMT (2564kb,D)

Link back to: arXiv, form interface, contact.