SAX-VSM: Interpretable Time Series Classification Using SAX and Vector Space Model

Senin, Pavel; Malinchik, Sergey

doi:10.1109/icdm.2013.52

Cited by 137 publications

(91 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The work of [9] proposes a variant of SAX approximation to produce distinct SAX-words with tf.idf weights (aka SAX-VSM) and builds a tf.idf centroid prototype for each class. This centroid-based approach was shown to produce results that are more accurate than the state-of-the-art, with an expensive training stage and a fast classification stage.…”

Section: Introductionmentioning

confidence: 99%

“…• SAX-VSEQL, can deal with X-axis offsets by learning variable-length symbolic words, therefore addressing a major weakness in prior work related to fixing the SAXword length across all time series [8], [9].…”

Section: Introductionmentioning

confidence: 99%

“…• SAX-VFSEQL, can deal with both X-axis and Y-axis offsets by learning fuzzy variable-length symbolic words, and removes the need for tuning SAX parameters, a task that is computationally expensive and was identified as an area in need of improvement by prior work [4], [9]. As shown in our experiments with UCR benchmark datasets, our approach removes the need for extensive tuning of SAX parameters, which results in fast training and test stages, while preserving high accuracy.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Time Series Classification by Sequence Learning in All-Subsequence Space

Nguyen¹,

Gsponer²,

Ifrim³

2017

2017 IEEE 33rd International Conference on Data Engineering (ICDE)

View full text Add to dashboard Cite

Abstract-Existing approaches to time series classification can be grouped into shape-based (numeric) and structure-based (symbolic). Shape-based techniques use the raw numeric time series with Euclidean or Dynamic Time Warping distance and a 1-Nearest Neighbor classifier. They are accurate, but computationally intensive. Structure-based methods discretize the raw data into symbolic representations, then extract features for classifiers. Recent symbolic methods have outperformed numeric ones regarding both accuracy and efficiency. Most approaches employ a bag-of-symbolic-words representation, but typically the word-length is fixed across all time series, an issue identified as a major weakness in the literature. Also, there are no prior attempts to use efficient sequence learning techniques to go beyond single words, to features based on variable-length sequences of words or symbols. We study an efficient linear classification approach, SEQL, originally designed for classification of symbolic sequences. SEQL learns discriminative subsequences from training data by exploiting the all-subsequence space using greedy gradient descent. We explore different discretization approaches, from none at all to increasing smoothing of the original data, and study the effect of these transformations on the accuracy of SEQL classifiers. We propose two adaptations of SEQL for time series data, SAX-VSEQL, can deal with X-axis offsets by learning variable-length symbolic words, and SAX-VFSEQL, can deal with X-axis and Y-axis offsets, by learning fuzzy variable-length symbolic words. Our models are linear classifiers in rich feature spaces. Their predictions are based on the most discriminative subsequences learned during training, and can be investigated for interpreting the classification decision.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Time Series Classification by Sequence Learning in All-Subsequence Space

Nguyen¹,

Gsponer²,

Ifrim³

2017

2017 IEEE 33rd International Conference on Data Engineering (ICDE)

View full text Add to dashboard Cite

show abstract

“…Many time series classification algorithms in the literature such as Shapelets [23,33], 1-NN BOSS [28] and SAX-VSM [29] have been shown to be competitive (and sometimes superior) to the state of the art, NN-DTW.…”

Section: Time Series Classificationmentioning

confidence: 99%

“…These modulations result in distortions of canonical temporal profiles that are well handled by DTW [20]. (2) Time series are too short for Bag-of-word-type approaches [28,29] to perform best. NN-DTW cannot scale to the typical size of satellite datasets where it is common to have 100 million example time series [9,10].…”

Section: Introductionmentioning

confidence: 99%

Indexing and classifying gigabytes of time series under time warping

Tan¹,

Webb²,

Petitjean³

2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

Time series classification maps time series to labels. The nearest neighbour algorithm (NN) using the Dynamic Time Warping (DTW) similarity measure is a leading algorithm for this task. NN compares each time series to be classified to every time series in the training database. With a training database of N time series of lengths L, each classification requires ϑ(N · L 2 ) computations. The databases used in almost all prior research have been relatively small (with less than 10, 000 samples) and much of the research has focused on making DTW's complexity linear with L, leading to a runtime complexity of O(N · L). As we demonstrate with an example in remote sensing, real-world time series databases are now reaching the million-to-billion scale. This wealth of training data brings the promise of higher accuracy, but raises a significant challenge because N is becoming the limiting factor. As DTW is not a metric, indexing objects induced by its space is extremely challenging. We tackle this task in this paper. We develop TSI, a novel algorithm for Time Series Indexing which combines a hierarchy of K-means clustering with DTW-based lower-bounding. We show that, on large databases, TSI makes it possible to classify time series orders of magnitude faster than the state of the art.

show abstract

Lagged encoding for image‐based time series classification using convolutional neural networks

Jastrzębska

2020

Statistical Analysis

View full text Add to dashboard Cite

Time series classification is a thriving area of research in machine learning. Among many applications, it is frequently applied to human activity analysis. Time series describing a human in motion are ubiquitously collected via omnipresent mobile devices and can be subjected to further processing. In this paper, we propose a novel, deep learning approach to time series classification. It is based on a lagged time series representation stored as images andConvolutional Neural Network used to image classification. We present a comparative study on different variants of lagged time series representation and we evaluate their effectiveness in a series of empirical experiments. We show that the developed method provides satisfying classification accuracy. The proposed image-based time series encoding is less resource-consuming than encodings used in other image-based approaches to time series classification. It is worth to emphasize that the proposed time series encoding conceals original time series values. Images are saved without scales and the order of observations cannot be reconstructed. Thus, the method is particularly suitable for systems that need to store sensitive information. K E Y W O R D S classification, convolutional neural network, image recognition, time series 1 Stat Anal Data Min: The ASA Data Sci Journal. 2020;13:245-260. wileyonlinelibrary.com/sam

show abstract

SAX-VSM: Interpretable Time Series Classification Using SAX and Vector Space Model

Cited by 137 publications

References 17 publications

Time Series Classification by Sequence Learning in All-Subsequence Space

Time Series Classification by Sequence Learning in All-Subsequence Space

Indexing and classifying gigabytes of time series under time warping

Lagged encoding for image‐based time series classification using convolutional neural networks

Contact Info

Product

Resources

About