Autoencoder-Based Iterative Modeling and Multivariate Time-Series Subsequence Clustering Algorithm (2022-09-09T00:00:00.000000Z)

TL;DR

An algorithm for the detection of change-points and the identification of the corresponding subsequences in transient multivariate time-series data (MTSD) is introduced that uses a recurrent neural network (RNN) based Autoencoder (AE) which is iteratively trained on incoming data.

Abstract

This paper introduces an algorithm for the detection of change-points and the identification of the corresponding subsequences in transient multivariate time-series data (MTSD). The analysis of such data has become increasingly important due to growing availability in many industrial fields. Labeling, sorting or filtering highly transient measurement data for training Condition-based Maintenance (CbM) models is cumbersome and error-prone. For some applications it can be sufficient to filter measurements by simple thresholds or finding change-points based on changes in mean value and variation. But a robust diagnosis of a component within a component group for example, which has a complex non-linear correlation between multiple sensor values, a simple approach would not be feasible. No meaningful and coherent measurement data, which could be used for training a CbM model, would emerge. Therefore, we introduce an algorithm that uses a recurrent neural network (RNN) based Autoencoder (AE) which is iteratively trained on incoming data. The scoring function uses the reconstruction error and latent space information. A model of the identified subsequence is saved and used for recognition of repeating subsequences as well as fast offline clustering. For evaluation, we propose a new similarity measure based on the curvature for a more intuitive time-series subsequence clustering metric. A comparison with seven other state-of-the-art algorithms and eight datasets shows the capability and the increased performance of our algorithm to cluster MTSD online and offline in conjunction with mechatronic systems.

Authors

Jonas Köhne

1 papers

Lars Henning

1 papers

C. Gühmann

2 papers

TL;DR

Abstract

Authors

References75 items

Time Series Segmentation Based on Stationarity Analysis to Improve New Samples Prediction

Aircraft Engine Run-to-Failure Dataset under Real Flight Conditions for Prognostics and Diagnostics

River: machine learning for streaming data in Python

Accurate Occupancy Detection of an Office Room From Light, Temperature, Humidity and CO2 Measurements Using Statistical Learning Models

ECHAD: Embedding-Based Change Detection From Multivariate Time Series in Smart Grids

Condition-Based Maintenance—An Extensive Literature Review

An Evaluation of Change Point Detection Algorithms

Unsupervised Anomaly Detection of Industrial Robots Using Sliding-Window Convolutional Variational Autoencoder

Selective review of offline change point detection methods

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Optimizing Data Stream Representation: An Extensive Survey on Stream Clustering Algorithms

Deep Clustering Based On A Mixture Of Autoencoders

Multidimensional Time Series Anomaly Detection: A GRU-based Gaussian Mixture Variational Autoencoder Approach

Clustering with Deep Learning: Taxonomy and New Methods

Time Series Segmentation through Automatic Feature Learning

Automatic feature extraction and selection for classification of cyclical time series data

Survey on time series motif discovery

State-of-the-art on clustering data streams

Unsupervised Interpretable Pattern Discovery in Time Series Using Autoencoders

Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clustering

A survey of methods for time series change point detection

Clustering Data Streams Based on Shared Density between Micro-Clusters

Accurate occupancy detection of an office room from light, temperature, humidity and CO2 measurements using statistical learning models

Time-series clustering - A decade review

A Comprehensive Survey of Clustering Algorithms

A Review of Subsequence Time Series Clustering

An empirical evaluation of similarity measures for time series classification

Data.gov

A database of C. elegans behavioral phenotypes

Spatio-Temporal Convolutional Sparse Auto-Encoder for Sequence Classification

SOStream: Self Organizing Density-Based Clustering over Data Stream

An effective evaluation measure for clustering on evolving data streams

Scikit-learn: Machine Learning in Python

Web-scale k-means clustering

Anomaly detection: A survey

Clustering high-dimensional data: A survey on subspace clustering, pattern-based clustering, and correlation clustering

Learning and Inferring Motion Patterns using Parametric Segmental Switching Linear Dynamic Systems

Bayesian Online Changepoint Detection

Efficient Learning of Sparse Representations with an Energy-Based Model

Characteristic-Based Clustering for Time Series Data

Pattern Recognition and Machine Learning

Clustering of time series data - a survey

A Framework for Clustering Evolving Data Streams

Fuzzy Clustering of Short Time-Series and Unevenly Distributed Sampling Points

Time series analysis.

Streaming-data algorithms for high-quality clustering

A Survey of Temporal Knowledge Discovery Paradigms and Methods

DETERMINISTIC CHAOS SEEN IN TERMS OF FEEDBACK CIRCUITS: ANALYSIS, SYNTHESIS, "LABYRINTH CHAOS"

Data clustering: a review

A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise

BIRCH: an efficient data clustering method for very large databases

Mining sequential patterns

Silhouettes: a graphical aid to the interpretation and validation of cluster analysis

A Cluster Separation Measure

Deterministic nonperiodic flow

A Novel Symmetric Stacked Autoencoder for Adversarial Domain Adaptation Under Variable Speed

MT3SCM: Multivariate Time Series Sub- Sequence Clustering Metric

ABIMCA: Autoencoder Based Iterative Modeling and Subsequence Clustering Algorithm. Python

Segmentation of Multivariate Industrial Time Series Data Based on Dynamic Latent Variable Predictability

2017).UCI Machine Learning Repository

Machine Learning and Data Mining in Pattern Recognition

ALGORITHMIC METHODS FOR SEGMENTATION OF TIME SERIES: AN OVERVIEW

Fault Diagnosis Systems

British Machine Vision Association

SERIES IN MACHINE PERCEPTION AND ARTIFICIAL INTELLIGENCE

Labyrinth Chaos

Density-Based Clustering over an Evolving Data Stream with Noise

Finding Motifs in Time Series

Segmenting Time Series: A Survey and Novel Approach

BIRCH”. In: ACM SIGMOD Record

Information Retrieval . London: Butterworths

A dendrite method for cluster analysis

c ○ 2002 Kluwer Academic Publishers. Manufactured in The Netherlands. Bayesian Clustering by Dynamics

Carnegie Mellon University-CMU Graphics Lab-Motion Capture Library

http:// mocap.cs.cmu.edu

Field of Study