Learn to Predict Vertical Track Irregularity with Extremely Imbalanced Data

Yutao Chen, Yu Zhang, Fei Yang
Proceedings of The 13th Asian Conference on Machine Learning, PMLR 157:1493-1504, 2021.

Abstract

Railway systems require regular manual maintenance, a large part of which is dedicated to inspecting track deformation. Such deformation might severely impact trains’ runtime security, whereas such inspections remain costly for both finance and human resources. Therefore, a more precise and efficient approach to detect railway track deformation is in urgent need. In this paper, we showcase an application framework for predicting vertical track irregularity, based on a real-world, large-scale dataset produced by several operating railways in China. We have conducted extensive experiments on various machine learning & ensemble learning algorithms in an effort to maximize the model’s capability in capturing any irregularity. We also proposed a novel approach for handling imbalanced data in multivariate time series prediction tasks with adaptive data sampling and penalized loss. Such an approach has proven to reduce models’ sensitivity to the imbalanced target domain, thus improving its performance in predicting rare extreme values.

Cite this Paper


BibTeX
@InProceedings{pmlr-v157-chen21c, title = {Learn to Predict Vertical Track Irregularity with Extremely Imbalanced Data}, author = {Chen, Yutao and Zhang, Yu and Yang, Fei}, booktitle = {Proceedings of The 13th Asian Conference on Machine Learning}, pages = {1493--1504}, year = {2021}, editor = {Balasubramanian, Vineeth N. and Tsang, Ivor}, volume = {157}, series = {Proceedings of Machine Learning Research}, month = {17--19 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v157/chen21c/chen21c.pdf}, url = {https://proceedings.mlr.press/v157/chen21c.html}, abstract = {Railway systems require regular manual maintenance, a large part of which is dedicated to inspecting track deformation. Such deformation might severely impact trains’ runtime security, whereas such inspections remain costly for both finance and human resources. Therefore, a more precise and efficient approach to detect railway track deformation is in urgent need. In this paper, we showcase an application framework for predicting vertical track irregularity, based on a real-world, large-scale dataset produced by several operating railways in China. We have conducted extensive experiments on various machine learning & ensemble learning algorithms in an effort to maximize the model’s capability in capturing any irregularity. We also proposed a novel approach for handling imbalanced data in multivariate time series prediction tasks with adaptive data sampling and penalized loss. Such an approach has proven to reduce models’ sensitivity to the imbalanced target domain, thus improving its performance in predicting rare extreme values.} }
Endnote
%0 Conference Paper %T Learn to Predict Vertical Track Irregularity with Extremely Imbalanced Data %A Yutao Chen %A Yu Zhang %A Fei Yang %B Proceedings of The 13th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Vineeth N. Balasubramanian %E Ivor Tsang %F pmlr-v157-chen21c %I PMLR %P 1493--1504 %U https://proceedings.mlr.press/v157/chen21c.html %V 157 %X Railway systems require regular manual maintenance, a large part of which is dedicated to inspecting track deformation. Such deformation might severely impact trains’ runtime security, whereas such inspections remain costly for both finance and human resources. Therefore, a more precise and efficient approach to detect railway track deformation is in urgent need. In this paper, we showcase an application framework for predicting vertical track irregularity, based on a real-world, large-scale dataset produced by several operating railways in China. We have conducted extensive experiments on various machine learning & ensemble learning algorithms in an effort to maximize the model’s capability in capturing any irregularity. We also proposed a novel approach for handling imbalanced data in multivariate time series prediction tasks with adaptive data sampling and penalized loss. Such an approach has proven to reduce models’ sensitivity to the imbalanced target domain, thus improving its performance in predicting rare extreme values.
APA
Chen, Y., Zhang, Y. & Yang, F.. (2021). Learn to Predict Vertical Track Irregularity with Extremely Imbalanced Data. Proceedings of The 13th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 157:1493-1504 Available from https://proceedings.mlr.press/v157/chen21c.html.

Related Material