Nearest-neighbor-based approach to time-series classification.

A very common and effective time series classification approach is the 1-Nearest Neighbor classier, with different distance measures such as the Euclidean or dynamic time warping distances. ... This paper provides an overview of current literature on time series classification approaches, in particular of early time series classification. ... Other time series classification approaches Many other approaches have been proposed to improve the strong performance and benefits of the 1-Nearest Neighbor algorithm outlined above. ...

doi:10.5281/zenodo.495150 fatcat:7f73zadtznblpdnuwmsdslf2pm

Open Access

In the last decade, the simple nearest neighbor classifier, in combination with dynamic time warping (DTW) as distance measure, has been shown to achieve surprisingly good overall results on time-series ... Many of the surveyed approaches were originally introduced for vector classification, and their application to time-series data is novel, therefore, we provide experimental results on large number of publicly ... The approach is based on the fuzzy k-nearest neighbor voting framework [27] . ...

doi:10.1007/978-3-662-45620-0_11 fatcat:q7dmxcs34jgf3hero47yjvp4kq

The result obtained by the Dynamic Time Warping is provided to the K-Nearest Neighbor Classifier to achieve Time Series Classification. ... Time series classification involves classifying the time series according to the labels given in the training dataset. Time series data has features that are not completely independent of each other. ... Also, the approach for Nearest Neighbor can be centroid-based or medoid-based to further reduce the computations. ACKNOWLEDGMENTS I would like to convey my heartfelt gratitude to my guide Prof. K. ...

doi:10.5120/ijca2017915974 fatcat:nb6az3mlxvejvcxjnlwbhrsdjq

Instance selection is a commonly applied approach for improving efficiency of nearest-neighbor classifier with respect to classification time. ... In this paper, we introduce a novel instance selection method that exploits the hubness phenomenon in time-series data, which states that some few instances tend to be much more frequently nearest neighbors ... Conclusion and Outlook We examined the problem of instance selection for speeding-up time-series classification. ...

doi:10.1007/978-3-642-20847-8_13 fatcat:6kgypdks7rfdzoi22pbpeic2by

The obtained segments refer to various types of human physical activity. To recognize these activities we use k-nearest neighbor algorithm and neural network as alternative. ... Primarily the method solves a problem of time series segmentation, assuming that each meaningful segment corresponds to one fundamental period of motion. ... To solve the time series classification problem in this paper we use the k-nearest neighbor method. The brief description of the method is provided below. ...

doi:10.1007/s11042-015-2643-0 fatcat:cuvfb3gkwzeelnoryxgsmgmtyi

Stateof-the-art solutions are based on machine learning. In this paper, we show that EEG datasets contain hubs, i.e., signals that appear as nearest neighbors of surprisingly many signals. ... Finally, we present the results of our empirical study on a large, publicly available collection of EEG signals and show that hubness-aware classifiers outperform the state-of-the-art time-series classifier ... Nenad Tomašev, researcher of the Artificial Intelligence Laboratory, Jožef Stefan Institute, Ljubljana, Slovenia as well as his contributions to the paper, esp. in Section 4.5 are greatly appreciated. ...

doi:10.12700/aph.13.2.2016.2.2 fatcat:m6u5v6jwirfrnga2nhygggz2wq

Open Access

A new, robust and computationally attractive approach to the problem of time series classification is discussed in this paper. ... Both the Bayesian as well as a new adaptive classification scheme for source selection are discussed. Simulation results are included to demonstrate the effectiveness of the new methodology. ... Adaptive time series classification The Bayesian approach The first approach to the problem discussed in the paper is the Bayesian one. ...

doi:10.1016/s0165-1684(96)00155-7 fatcat:yokouc5r4bcr5kffu2owwrmxqe

In our research, however, we observed that a single fixed choice for the number of nearest neighbors k may lead to suboptimal performance. ... This refers to estimating the expected classification accuracy for each time series and each k individually. ... As we propose the classification of time-series based on a quality score estimated individually for each of them, the proposed approach is called time-series classification based on individual quality ...

doi:10.1109/cidm.2011.5949441 dblp:conf/cidm/BuzaNS11 fatcat:etcf33blibdhzkdou74ztnmpni

This paper describes solution for Time Series Land Cover Classification Challenge (TiSeLaC). ... Using features extracted from satellite images time series (SITS) each pixel corresponding to 30m⇥30m area can be classified to one of general class (urban area, forest, water, etc.). ... k-NN model Here the non-parametric classification is applied to predict class based on nearest neighbors using coordinates provided. ...

dblp:conf/pkdd/Sergey17 fatcat:x4pc5gwoivbjlb4ojnoojbwi3q

We formulate the task as a time-series classification problem, point out that state-of-the-art solutions are capable to solve this problem with a high accuracy. ... In our experiments on publicly available real ECG-data, we empirically evaluate our approach and show that it outperforms a state-of-the-art instance selection technique. ... Attempts to speed up DTW-based nearest neighbor (NN) classification fall into 4 major categories: i) speed-up the calculation of the distance of two time series, ii) reduce the length of time series, iii ...

doi:10.1109/hisb.2011.26 dblp:conf/hisb/BuzaNSK11 fatcat:izxh43ilcrd7rfwfowfpjipubu

SFM is inspired by the optimization model of support vector machine and the nearest neighbor rule to incorporate both spatial and temporal of the multi-dimensional time series data. ... In this study, a novel multidimensional time series classification technique, namely support feature machine (SFM), is proposed. ... The third step is the classification algorithm, which is based on optimization models used to classify unlabeled samples based on the nearest neighbor rule. ...

doi:10.1145/1281192.1281208 dblp:conf/kdd/ChaovalitwongseFS07 fatcat:2xxndlvi4zdjddsi2mq7szebuu

The problem of time series classification has attracted great interest in the last decade. However current research assumes the existence of large amounts of labeled training data. ... While such algorithms are well known in text domains, we will show that special considerations must be made to make them both efficient and effective for the time series domain. ... Our point is simply that if you want accurate classification of time series, one-nearest-neighbor with Euclidean distance is very hard to beat. ...

doi:10.1145/1150402.1150498 dblp:conf/kdd/WeiK06 fatcat:3jxz5g4oyjhgnc3wfe7qlmyxrq

Many algorithms have been proposed for the problem of time series classification. ... However, it is clear that one-nearest-neighbor with Dynamic Time Warping (DTW) distance is exceptionally difficult to beat. ... In Rodriguez & Alonso (2004) , the authors use a DTW based decision tree to classify time series. ...

doi:10.1145/1143844.1143974 dblp:conf/icml/XiKSWR06 fatcat:frkt5c3ezffydfor4sldonlpfm

For example, modern smartphones have at least a dozen sensors capable of producing streaming time series, and hospital-based (and increasingly, home-based) medical devices can produce time series streams ... In this work, we introduce a novel framework for multi-dimensional time series classification that weights the class prediction from each time series stream. ... ACKNOWLEDGMENTS We would like to acknowledge the financial support for our research provided by NSF IIS-1161997II. We thank all the donors of datasets. ...

doi:10.1109/icdm.2013.33 dblp:conf/icdm/00010ZUK13 fatcat:yjhoh33jwjehbf2hs65uijmi54

In this paper a new hybrid approach to learn a global constraint of DTW distance is proposed. This approach is based on Large Margin Nearest Neighbors classification and Tabu Search algorithm. ... Experiments show the effectiveness of this approach to improve time series classification results. ... Large Margin Nearest Neighbors Classification In the proposed approach to learning a global constraint DTW, the wide margin nearest 'neighbors algorithm (LMNN) is used. ...

doi:10.1016/j.procs.2016.04.003 fatcat:ckmohefo3vfadjhv7z5ogvcea4

Open Access

A Literature Survey Of Early Time Series Classification And Deep Learning

Preserved Fulltext

Hubness-Aware Classification, Instance Selection and Feature Construction: Survey and Extensions to Time-Series [chapter]

Preserved Fulltext

Effect of Dynamic Time Warping using different Distance Measures on Time Series Classification

Preserved Fulltext

INSIGHT: Efficient and Effective Instance Selection for Time-Series Classification [chapter]

Preserved Fulltext

Human activity recognition using quasiperiodic time series collected from a single tri-axial accelerometer

Preserved Fulltext

Classification of Electroencephalograph Data: A Hubness-aware Approach

Preserved Fulltext

A new time series classification approach

Preserved Fulltext

IQ estimation for accurate time-series classification

Preserved Fulltext

Temporal and spatial approaches for land cover classification

Preserved Fulltext

Fast Classification of Electrocardiograph Signals via Instance Selection

Preserved Fulltext

Support feature machine for classification of abnormal brain activity

Preserved Fulltext

Semi-supervised time series classification

Preserved Fulltext

Fast time series classification using numerosity reduction

Preserved Fulltext

Classification of Multi-dimensional Streaming Time Series by Weighting Each Classifier's Track Record

Preserved Fulltext

DTW-Global Constraint Learning Using Tabu Search Algorithm

Preserved Fulltext