Improving the Accuracy of Global Forecasting Models using Time Series Data Augmentation

@article{Bandara2020ImprovingTA,
  title={Improving the Accuracy of Global Forecasting Models using Time Series Data Augmentation},
  author={Kasun Bandara and Hansika Hewamalage and Yuan-Hao Liu and Yanfei Kang and C. Bergmeir},
  journal={Pattern Recognit.},
  year={2020},
  volume={120},
  pages={108148}
}

Figures and Tables from this paper

Exploring Generative Data Augmentation in Multivariate Time Series Forecasting : Opportunities and Challenges

This work introduces a novel meta-algorithm for time-series data augmentation to address the data scarcity problem, and generates synthetic MTS data preserving temporal dynamics using an offthe-shelf generative algorithm and frame augmentation in MTS as a transfer learning problem.

Time Series Data Augmentation for Deep Learning: A Survey

This paper systematically review different data augmentation methods for time series, and proposes a taxonomy for the reviewed methods, and provides a structured review for these methods by highlighting their strengths and limitations.

Data Augmentation techniques in time series domain: A survey and taxonomy

This study systematically review the current state-of-the-art in the area of deep learning generative models to provide an overview of all available algorithms and proposes a taxonomy of the most relevant researches.

Robust Augmentation for Multivariate Time Series Classification

This paper shows that the simple methods of cutout, cutmix, mixup, and window warp improve the robustness and overall performance in a statistically significant way for convolutional, recurrent, and self-attention based architectures for time series classification.

Financial Time Series Data Augmentation with Generative Adversarial Networks and Extended Intertemporal Return Plots

State-of-the art image-based generative models are applied for the task of data augmentation and the extended intertemporal return plot (XIRP), a new image representation for time series are introduced and effective in reducing the return forecast error.

Time series data augmentation method of small sample based on optimized generative adversarial network

The experimental results on the UCR time series dataset show that the long short‐term memory network classification accuracy has been effectively improved after using the GC‐GAN model, with a maximum improvement of 4.4%, indicating that the method proposed in this article has better data augmentation capabilities.

Application of Neural Networks in Financial Time Series Forecasting Models

  • Xinhui Li
  • Economics
    Journal of Function Spaces
  • 2022
At present, the economic development of the world’s major economies is showing a positive and positive state. Driven by the development of related industries, the development of the financial field

The Wisdom of the Data: Getting the Most Out of Univariate Time Series Forecasting

This article presents the concept of the “wisdom of the data” and how data manipulation can result in information extraction which translates to improved forecast accuracy by aggregating (combining) forecasts computed on different perspectives of the same data.

References

SHOWING 1-10 OF 84 REFERENCES

Conditional Time Series Forecasting with Convolutional Neural Networks

This paper compares the performance of the WaveNet model to a state-of-the-art fully convolutional network (FCN), and an autoregressive model popular in econometrics and shows that the model is much better able to learn important dependencies in between financial time series resulting in a more robust and accurate forecast.

A novel transfer learning framework for time series forecasting

LSTM-MSNet: Leveraging Forecasts on Sets of Related Time Series With Multiple Seasonal Patterns

This article proposes long short-term memory multiseasonal net (LSTM-MSNet), a decomposition-based unified prediction framework to forecast time series with multiple seasonal patterns, outperforming many state-of-the-art multise Masonal forecasting methods.

DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks

Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networks

A novel deep learning framework, namely Long- and Short-term Time-series network (LSTNet), to address this open challenge of multivariate time series forecasting, using the Convolution Neural Network and the Recurrent Neural Network to extract short-term local dependency patterns among variables and to discover long-term patterns for time series trends.

GRATIS: GeneRAting TIme Series with diverse and controllable characteristics

This work proposes GeneRAting TIme Series with diverse and controllable characteristics, named GRATIS, with the use of mixture autoregressive (MAR) models, and demonstrates the usefulness of the time series generation process through a time series forecasting application.

Recurrent Neural Networks for Time Series Forecasting: Current Status and Future Directions

Generating Synthetic Time Series to Augment Sparse Datasets

This paper proposes new data augmentation techniques specifically designed for time series classification, where the space in which they are embedded is induced by Dynamic Time Warping (DTW).

Probabilistic Forecasting with Temporal Convolutional Neural Network

...