Imaging Time-Series to Improve Classification and Imputation

By Zhiguang Wang, Tim Oates. International Joint Conference on Artificial Intelligence, 2015

928 citations recorded by the source corpus.

Inspired by recent successes of deep learning in computer vision, we propose a novel framework for encoding time series as different types of images, namely, Gramian Angular Summation/Difference Fields (GASF/GADF) and Markov Transition Fields (MTF). This enables the use of techniques from computer vision for time series classification and imputation. We used Tiled Convolutional Neural Networks (tiled CNNs) on 20 standard datasets to learn high-level features from the individual and compound GASF-GADF-MTF images. Our approaches achieve highly competitive results when compared to nine of the current best time series classification approaches. Inspired by the bijection property of GASF on 0/1 rescaled data, we train Denoised Auto-encoders (DA) on the GASF images of four standard and one synthesized compound dataset. The imputation MSE on test data is reduced by 12.18%-48.02% when compared to using the raw data. An analysis of the features and weights learned via tiled CNNs and DAs explains why the approaches work.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/69eea40aa3eed818485665926808b20328054279

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jindong Wang
2. Rahul Katarya
3. Anmol Mahajan
4. Tao Qin
5. Wenjie Feng
6. Weizhong Yan
7. Tim Oates
8. Zhiguang Wang

In Signal Processing

1. Rahul Katarya
2. Anmol Mahajan
3. Weizhong Yan
4. Jindong Wang
5. Tim Oates
6. Zhiguang Wang
7. Tao Qin
8. Sinno Jialin Pan

Citations in the corpus

• Time series classification from scratch with deep neural networks: A strong baseline (2016, methodology): “We adopt the Gramian Angular Summation Field (GASF) [21] to visualize the filters/weights in the neural networks.”
• A survey of neural network techniques in market trend analysis (2017, methodology): “GAF enables the use of techniques from computer vision for time series classification[17].”
• Financial Time-Series Data Analysis Using Deep Convolutional Neural Networks (2016, positive): “According to [30], there are several advantages of GAF representation method: the bottom right element in the matrix contained the last information in the time series.”
• AdaRNN: Adaptive Learning and Forecasting of Time Series (2021, methodology): “Cosine distance is widely used in time series data [25, 50] to measure the similarity between two distributions.”