A review on distance based time series classification

Abanda, Amaia; Mori, Usue; Lozano, José A.

doi:10.1007/s10618-018-0596-4

Cited by 209 publications

(116 citation statements)

References 95 publications

(130 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The classification task is done by metric classifiers, such as Support Vector Machines (SVM) and K-Nearest Neighbors (KNN). SVM classifiers have been using embedded distance features for time-series data by computing EDs inside the classification task [6]. The SVM method is an effective approach in case of high dimensional data and superior in classification accuracy [7].…”

Section: Introductionmentioning

confidence: 99%

Quantitative and Qualitative Analysis of Time-Series Classification Using Deep Learning

et al. 2020

View full text Add to dashboard Cite

Time-series classification is utilized in a variety of applications leading to the development of many data mining techniques for time-series analysis. Among the broad range of time-series classification algorithms, recent studies are considering the impact of deep learning methods on time-series classification tasks. The quantity of related publications requires a bibliometric study to explore most prominent keywords, countries, sources and research clusters. The paper conducts a bibliometric analysis on related publications in time-series classification, adopted from Scopus database between 2010 and 2019. Through keywords cooccurrence analysis, a visual network structure of top keywords in time-series classification research has been produced and deep learning has been introduced as the most common topic by additional inquiry of the bibliography. The paper continues by exploring the publication trends of recent deep learning approaches for time-series classification. The annual number of publications, the productive and collaborative countries, the growth rate of sources, the most occurred keywords and the research collaborations are revealed from the bibliometric analysis within the study period. The research field has been broken down into three main categories as different frameworks of deep neural networks, different applications in remote sensing and also in signal processing for time-series classification tasks. The qualitative analysis highlights the categories of top citation rate papers by describing them in details. INDEX TERMS Time-series classification, deep learning, remote sensing, signal processing, bibliometrics, research productivity.

show abstract

Section: Introductionmentioning

confidence: 99%

Quantitative and Qualitative Analysis of Time-Series Classification Using Deep Learning

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The choice of this classifier over other possibilities was supported by the results of [15]. In future applications, the current approach could be extended by implementing other time series classifiers and/or measures of proximity in the PREF-Constructor algorithm [34].…”

Section: Plos Onementioning

confidence: 96%

Assessing site signal preservation in reference chronologies for dendro-provenancing

Gut

2020

PLoS ONE

View full text Add to dashboard Cite

Regional differences in tree growth can be used to approximate the geographical provenance of ring-width series ('dendro-provenancing'). This method relies on cross-dated ringwidth series (reference chronologies) that are thought to represent the radial growth signal of trees in a given region. Reference chronologies are often established from ring-width series of living tree populations. Frequently, they are too short to allow for investigating the provenance of historical wood. Thus, references are extended by ring-width series from buildings and art-historical objects that exhibit best matching growth patterns with the living tree references. Yet, series from other provenances may erroneously be included. Thereby the local or regional growth signal of the references is progressively contaminated, but this has received little attention to date. I investigate this contamination risk using a simulation approach that allows for generating pseudo site chronologies that preserve the relevant statistical properties of the real site chronologies. While the exact provenance of historical wood is unknown, for simulated ring-width series the provenance is unambiguous. Hence, pseudo reference chronologies may be established while monitoring the signal mixture. Specifically, 15 site chronologies of Norway spruce (Picea abies (L.) H. Karst.) from northeastern Switzerland were used to generate 15 pseudo site growth signals that span 1000 years. The simulation demonstrates that quasi uncontaminated references can be established in ideal circumstances for the study area. However, the thresholds for the similarity in between-series correlation must be very high. Even then, contaminated pseudo references occurred in rare cases during the simulation. Yet, elevation-specific pseudo references were established with lower thresholds. Simulation currently offers the only approach for assessing the contamination risk of reference chronologies, and it allows for elucidating the conditions under which acceptable levels of contamination can be guaranteed. Therefore, the present approach paves the way towards a practical simulation tool for dendroprovenancing.

show abstract

“…Lines and Bagnall [33] defines an ensemble classifier of elastic distance measures, which significantly outperforms the individual classifiers with one kind of distance measure. Furthermore, Abanda et al [1] gives a comprehensive review of distance-based time series classification.…”

Section: Distance-based Time Series Classificationmentioning

confidence: 99%

Time series classification: nearest neighbor versus deep learning models

Jiang

2020

SN Appl. Sci.

View full text Add to dashboard Cite

Time series classification has been an important and challenging research task. In different domains, time series show different patterns, which makes it difficult to design a global optimal solution and requires a comprehensive evaluation of different classifiers across multiple datasets. With the rise of big data and cloud computing, deep learning models, especially deep neural networks, arise as a new paradigm for many problems, including image classification, object detection and natural language processing. In recent years, deep learning models are also applied for time series classification and show superiority over traditional models. However, the previous evaluation is usually limited to a small number of datasets and lack of significance analysis. In this study, we give a comprehensive comparison between nearest neighbor and deep learning models. Specifically, we compare 1-NN classifiers with eight different distance measures and three state-of-the-art deep learning models on 128 time series datasets. Our results indicate that deep learning models are not significantly better than 1-NN classifiers with edit distance with real penalty and dynamic time warping.

show abstract

A review on distance based time series classification

Cited by 209 publications

References 95 publications

Quantitative and Qualitative Analysis of Time-Series Classification Using Deep Learning

Quantitative and Qualitative Analysis of Time-Series Classification Using Deep Learning

Assessing site signal preservation in reference chronologies for dendro-provenancing

Time series classification: nearest neighbor versus deep learning models

Contact Info

Product

Resources

About