Fast Channel Selection for Scalable Multivariate Time Series Classification

Dhariyal, Bhaskar; Nguyen, Thach Le; Ifrim, Georgiana

doi:10.1007/978-3-030-91445-5_3

Cited by 8 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This study extends the work done in (Dhariyal et al, 2021) on fast channel selection for multivariate time series classification. Our focus here is on extending the channel selection methodology, with further attention to the individual building blocks and ways to increase the robustness to noise in the input data.…”

Section: Introductionsupporting

confidence: 71%

“…In order to compare our methods to a strong baseline, we adapt the filter feature selection method from scikit-klearn 2 to work with multivariate time series, by considering each channel as a single feature. In this work we extend our preliminary study (Dhariyal et al, 2021) with further methods and evaluation, as well as provide all data and code for reproducing the results 3 .…”

Section: Channel Selection For Multivariate Time Series Classificationmentioning

confidence: 93%

“…In our approach, it is also important that the class prototypes are inexpensive to compute. Extending our previous work (Dhariyal et al, 2021) on channel selection, here we investigate other choices for the class prototypes, with a specific focus on robustness to noise. We focus on three prototypes: mean, median and MAD (data trimmed using median absolute deviation (Leys et al, 2013)).…”

Section: Class Prototypesmentioning

confidence: 99%

See 2 more Smart Citations

Scalable Classifier-Agnostic Channel Selection for Multivariate Time Series Classification

Dhariyal¹,

Nguyen²,

Ifrim³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Accuracy is a key focus of current work in time series classification. However, in many applications speed and data reduction is equally important, especially when the data scale and storage requirements increase rapidly. Current multivariate time series classification algorithms need hundreds of compute hours to complete training and prediction. This is due to the nature of multivariate time series data which grows with the number of time series, their length and the number of channels. In many applications not all the channels are useful for the classification task, hence we require methods that can efficiently select useful channels and thus save computational resources. We propose and evaluate two methods for channel selection. Our techniques work by representing each class by a prototype time series and performing channel selection based on the prototype-distance between classes. The main hypothesis is that useful channels enable better separation between classes and hence channels with higher distance between class-prototypes are more useful. On the UEA Multivariate Time Series Classification (MTSC) benchmark we show that these techniques achieve significant data reduction and classifier speedup, for similar levels of classification accuracy. Channel selection is applied as a pre-processing step before training state-of-the-art MTSC algorithms and saves about 70% of computation time and data storage, with preserved accuracy. Furthermore, our methods enable even efficient classifiers, such as ROCKET, to achieve better accuracy as compared to using no channel selection or forward channel selection. To further study the impact of our techniques we present experiments on classifying synthetic multivariate time series datasets with more than 100 channels, as well as a real-world case study on a dataset with 50 channels. We find that our channel selection methods lead to significant data reduction with preserved or improved accuracy.

show abstract

Section: Introductionsupporting

confidence: 71%

Section: Channel Selection For Multivariate Time Series Classificationmentioning

confidence: 93%

Section: Class Prototypesmentioning

confidence: 99%

See 1 more Smart Citation

Scalable Classifier-Agnostic Channel Selection for Multivariate Time Series Classification

Dhariyal¹,

Nguyen²,

Ifrim³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, when working with complex multivariate time series datasets, this can be a challenging task due to the dimensions of each instance. To simplify the process of creating the metafeature dataset, I employed the random sampling method as in [2] and key channel selection methods described in [5]. The random sampling method reduces the number of instances in each dataset, while the key channel selection method reduces the number of dimensions in each instance.…”

Section: Deep-learning Classi Er Recommendation In Dl-mtscrmentioning

confidence: 99%

“…The subsample ratio was consistent with [2] and is displayed in Table 2. For the key channel selection algorithm, I utilized Elbow Class Pairwise (ECP) [22], which is implemented in sktime, a Python library for time series analysis. The training outcome pertaining to the reduced training datasets served as the meta-attribute.…”

Section: Experimental Set-upmentioning

confidence: 99%

Suitable Deep Learning Classifier Recommendation for Multi-variate Time Series Classification

Gan

2024

Preprint

View full text Add to dashboard Cite

Time series classification can be categorized into two main types: univariate and multivariate. The major difference between the two is that multivariate datasets have multiple dimensions, which makes them more intricate than univariate datasets. Selecting a suitable classifier for multivariate time series classification can be challenging, as it involves trying out several algorithms to handle the intricate data. While some studies have been conducted on automatically finding suitable classifiers for univariate datasets, there are currently no such studies for multivariate ones. To fill this gap, this paper introduces a recommendation system for deep-learning classifiers that are suitable for multivariate time series classification. The experiments demonstrate that this recommendation system can automatically identify the suitable deep-learning classifier for some multivariate time series datasets.

show abstract

Dimension Selection Strategies for Multivariate Time Series Classification with HIVE-COTEv2.0

Ruiz

Bagnall

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Fast Channel Selection for Scalable Multivariate Time Series Classification

Cited by 8 publications

References 14 publications

Scalable Classifier-Agnostic Channel Selection for Multivariate Time Series Classification

Scalable Classifier-Agnostic Channel Selection for Multivariate Time Series Classification

Suitable Deep Learning Classifier Recommendation for Multi-variate Time Series Classification

Dimension Selection Strategies for Multivariate Time Series Classification with HIVE-COTEv2.0

Contact Info

Product

Resources

About