Transfer Learning With Time Series Data: A Systematic Mapping Study

被引：20

作者：

Weber, Manuel ^{[1
]}

Auch, Maximilian ^{[1
]}

Doblander, Christoph ^{[2
]}

Mandl, Peter ^{[1
]}

Jacobsen, Hans-Arno ^{[3
]}

机构：

[1] Munich Univ Appl Sci HM, Dept Comp Sci & Math, D-80335 Munich, Germany

[2] Tech Univ Munich TUM, Chair Applicat & Middleware Syst, D-85748 Garching, Germany

[3] Univ Toronto, Dept Elect & Comp Engn, Toronto, ON M5S 3G4, Canada

来源：

IEEE ACCESS | 2021年 / 9卷

关键词：

Time series analysis; Deep learning; Transfer learning; Systematics; Forecasting; Computational modeling; Predictive models; Time series; transfer learning; domain adaptation; deep learning; survey; FAULT-DIAGNOSIS APPROACH; DOMAIN ADAPTATION; ACTIVITY RECOGNITION; CLASSIFICATION; AUTOENCODER; PREDICTION; SPEECH; KNOWLEDGE; FRAMEWORK; MODEL;

D O I：

10.1109/ACCESS.2021.3134628

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Transfer Learning is a well-studied concept in machine learning, that relaxes the assumption that training and testing data need to be drawn from the same distribution. Recent success in applying transfer learning in the area of computer vision has motivated research on transfer learning also in context of time series data. This benefits learning in various time series domains, including a variety of domains based on sensor values. In this paper, we conduct a systematic mapping study of literature on transfer learning with time series data. Following the review guidelines of Kitchenham and Charters, we identify and analyze 223 relevant publications. We describe the pursued approaches and point out trends. Especially during the last two years, there has been a vast increase in the number of publications on the topic. This paper's findings can help researchers as well as practitioners getting into the field and can help identify research gaps.

引用

页码：165409 / 165432

页数：24

共 168 条

[1] Transferring Activity Recognition Models for New Wearable Sensors with Deep Generative Domain Adaptation [J].

Akbari, Ali ;

Jafari, Roozbeh .

IPSN '19: PROCEEDINGS OF THE 2019 INTERNATIONAL CONFERENCE ON INFORMATION PROCESSING IN SENSOR NETWORKS, 2019, :85-96

[2]

Al Hafiz Khan Md Abdullah, 2017, 2017 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops), P545, DOI 10.1109/PERCOMW.2017.7917621

[3]

Alves T, 2018, IEEE INT CONF BIG DA, P1328, DOI 10.1109/BigData.2018.8621927

[4]

[Anonymous], 2008, A literature survey on domain adaptation of statistical classifiers

[5] A Scalable Room Occupancy Prediction with Transferable Time Series Decomposition of CO2 Sensor Data [J].

Arief-Ang, Irvan B. ;

Hamilton, Margaret ;

Salim, Flora D. .

ACM TRANSACTIONS ON SENSOR NETWORKS, 2018, 14 (3-4)

[6] The great time series classification bake off: a review and experimental evaluation of recent algorithmic advances [J].

Bagnall, Anthony ;

Lines, Jason ;

Bostrom, Aaron ;

Large, James ;

Keogh, Eamonn .

DATA MINING AND KNOWLEDGE DISCOVERY, 2017, 31 (03) :606-660

[7] Transfer Learning for Leisure Centre Energy Consumption Prediction [J].

Banda, Paul ;

Bhuiyan, Muhammed A. ;

Zhang, Kevin ;

Song, Andy .

COMPUTATIONAL SCIENCE - ICCS 2019, PT I, 2019, 11536 :112-123

[8] A Deep Transfer Learning Approach for Improved Post-Traumatic Stress Disorder Diagnosis [J].

Banerjee, Debrup ;

Islam, Kazi ;

Mei, Gang ;

Xiao, Lemin ;

Zhang, Guangfan ;

Xu, Roger ;

Ji, Shuiwang ;

Li, Jiang .

2017 17TH IEEE INTERNATIONAL CONFERENCE ON DATA MINING (ICDM), 2017, :11-20

[9]

Benchaira K, 2019, P 4 INT C BIG DAT IN, P1

[10]

Blanke U., 2010, P INT S WEAR COMP IS, P1

← 1 2 3 4 5 6 7 8 9 10 →