2021年A类B类会议时间序列相关研究论文汇总
前些天学姐的学习交流群里面有同学提出学姐推的论文都太老了！可是前几天推的都是经典论文~肯定是旧一点的，但是新的论文也是学习了旧的论文发现了新的研究点写出来的，所以经典论文不可丢。
今天学姐整理了ACM CIKM、ICDE、ICML、KDD、IJCAI等期刊会议的时间序列2021年的论文给大家。记得收藏点赞哦！
CIKM
ACM CIKM（International Conference on Information and Knowledge Management）是信息知识管理和数据挖掘领域顶级学术会议之一，其目的是确定未来知识和信息系统发展面临的具有挑战性的问题，并通过征求和审查高质量的应用和理论研究成果来塑造未来的研究方向。被中国计算机协会CCF推荐为B类会议。
ACM CIKM 2021主要包含7个tracks，其中关注的三个tracks对应的接收率为：
Full Paper Track（271/1251=21.7%），
Applied Paper Track（70/290=24.1%），
Short Paper Track（177/626=28.3%）。
下面整理了以上三个Tracks中时间序列相关的研究成果。主要研究方向包括：时间序列预测&异常检测，时间序列分割，时间序列分类 及其交叉应用领域。
01
论文标题：
ClaSP - Time Series Segmentation
论文地址：
https://dl.acm.org/doi/10.1145/3459637.3482240
源码链接：
https://github.com/alan-turing-institute/sktime
02
论文标题
Learning to Learn the Future: Modeling Concept Drifts in Time Series Prediction
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482271
源码链接
暂未开源
03
论文标题
AdaRNN: Adaptive Learning and Forecasting of Time Series
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482315
源码链接
https://github.com/jindongwang/transferlearning/tree/master/code/deep/adarnn
04
论文标题
Learning Saliency Maps to Explain Deep Time Series Classifiers
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482446
源码链接
https://github.com/kingspp/timeseries-explain
05
论文标题
Actionable Insights in Urban Multivariate Time-series
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482410
源码链接
https://github.com/AdityaLab/RaTSS
06
论文标题
Integrating Static and Time-Series Data in Deep Recurrent Models for Oncology Early Warning Systems
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482441
源码链接
暂未开源
07
论文标题
Historical Inertia: A Neglected but Powerful Baseline for Long Sequence Time-series Forecasting
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482120
源码链接
https://github.com/TakuyaShintate/tsts
08
论文标题
Discovering Time-invariant Causal Structure from Temporal Data
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482086
源码链接
暂未开源
09
论文标题
AGCNT: Adaptive Graph Convolutional Network for Transformer-based Long Sequence Time-Series Forecasting
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482054
源码链接
暂未开源
10
论文标题
Improving Irregularly Sampled Time Series Learning with Time-Aware Dual-Attention Memory-Augmented Networks
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482079
源码链接
暂未开源
11
论文标题
BiCMTS: Bidirectional Coupled Multivariate Learning of Irregular Time Series with Missing Values
论文地址
https://dl.acm.org/doi/10.1145/3459637.3482064
源码链接
暂未开源
ICDE
IEEE ICDE（IEEE International Conference on Data Engineering）是 IEEE 的旗舰会议，和 SIGMOD、VLDB 并称数据库领域三大顶会，旨在解决设计、构建、管理和评估高级数据密集型系统和应用程序中的研究问题，是研究人员、从业人员、开发人员和用户探索前沿思想并交流技术、工具和经验的领先论坛。被中国计算机协会CCF推荐为 A类会议。
IEEE ICDE 会议的多年平均接受率为 19.1%， 2021论文接收率暂无。
下面整理的是ICDE 2021 Research Track (Full & Short) & TKDE Posters中时间序列相关研究成果。
01
论文标题
EnhanceNet: Plugin Neural Networks for Enhancing Correlated Time Series Forecasting
论文地址
https://ieeexplore.ieee.org/document/9458855
源码链接
https://github.com/alan-turing-institute/sktime
02
论文标题
Scalable Model-Based Management of Correlated Dimensional Time Series in ModelarDB+
论文地址
https://ieeexplore.ieee.org/document/9458830
源码链接
https://github.com/skejserjensen/ModelarDB
03
论文标题
TS-Benchmark: A Benchmark for Time Series Databases
论文地址
https://ieeexplore.ieee.org/document/9458659
源码链接
https://github.com/timescale/tsbs
04
论文标题
DAEMON: Unsupervised Anomaly Detection and Interpretation for Multivariate Time Series
论文地址
https://ieeexplore.ieee.org/document/9458835
源码链接
暂无
05
论文标题
GRAB: Finding Time Series Natural Structures via A Novel Graph-based Scheme
论文地址
https://ieeexplore.ieee.org/document/9458905
源码链接
暂无
06
论文标题
An Actor-Critic Ensemble Aggregation Model for Time-Series Forecasting
论文地址
https://ieeexplore.ieee.org/document/9458798
源码链接
暂无
07
论文标题
Efficient Shapelet Discovery for Time Series Classification
论文地址
https://ieeexplore.ieee.org/document/9458803
源码链接
暂无
想要源码的可以找原作者，给原作者发邮件申请嗷！
ICML
ICML（International Conference on Machine Learning, ICML）是由国际机器学习学会（IMLS）主办的年度机器学习国际顶级会议，是机器学习领域发展的重要会议，且被中国计算机协会CCF推荐为 A类会议。
ICML 2021 的论文接收结果：共有5513篇有效投稿，其中1184篇论文被接收，接收率为21.4%。在被接收论文中，有166篇长文和1018篇短文。
以下整理了ICML 2021 Long Representation中时间序列相关研究成果。主要方向包括不局限于时序预测、时序异常检测、时序分类、时序因果分析、多元时序等。
Papers
01
论文标题
Voice2Series: Reprogramming Acoustic Models for Time Series Classification
论文地址
http://proceedings.mlr.press/v139/yang21j.html
源码链接
https://github.com/huckiyang/Voice2Series-Reprogramming
02
论文标题
Neural Rough Differential Equations for Long Time Series
论文地址
http://proceedings.mlr.press/v139/morrill21b.html
源码链接
https://github.com/jambo6/neuralRDEs
03
论文标题
Necessary and sufficient conditions for causal feature selection in time series with latent common causes
论文地址
http://proceedings.mlr.press/v139/mastakouri21a.html
源码链接
暂未开源
04
论文标题
Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting
论文地址
http://proceedings.mlr.press/v139/rasul21a.html
源码链接
暂未开源
05
论文标题
Conformal prediction interval for dynamic time-series
论文地址
http://proceedings.mlr.press/v139/xu21h.html
源码链接
https://github.com/hamrel-cxu/EnbPI
06
论文标题
Z-GCNETs: Time Zigzags at Graph Convolutional Networks for Time Series Forecasting
论文地址
http://proceedings.mlr.press/v139/chen21o.html
源码链接
https://github.com/Z-GCNETs/Z-GCNETs
07
论文标题
End-to-End Learning of Coherent Probabilistic Forecasts for Hierarchical Time Series
论文地址
http://proceedings.mlr.press/v139/rangapuram21a.html
源码链接
https://github.com/awslabs/gluon-ts
08
论文标题
Approximation Theory of Convolutional Architectures for Time Series Modelling
论文地址
http://proceedings.mlr.press/v139/jiang21d.html
源码链接
暂未开源
09
论文标题
Whittle Networks: A Deep Likelihood Model for Time Series
论文地址
http://proceedings.mlr.press/v139/yu21c.html
源码链接
https://github.com/ml-research/WhittleNetworks
10
论文标题
Explaining Time Series Predictions with Dynamic Masks
论文地址
http://proceedings.mlr.press/v139/crabbe21a.html
源码链接
https://github.com/JonathanCrabbe/Dynamask
KDD
ACM SIGKDD（Conference on Knowledge Discovery and Data Mining, KDD）是全球最大规模的国际数据科学会议，将展示知识发现和数据挖掘方面的最新研究成果。SIG是ACM中关于知识发现和数据挖掘的特别兴趣小组，KDD知识发现和数据挖掘年度国际会议是数据挖掘、数据科学和分析领域的首屈一指的跨学科会议，被中国计算机协会CCF推荐为A类会议。
KDD 2021 的论文接收结果：共有1541篇有效投稿，其中238篇论文被接收，接收率为15.44%。相比 KDD 2020 的接收率 16.9% (216/1279)有所下降。
以下主要整理了KDD 2021 Research Track & Applied Data Science Track中时间序列相关研究成果。
01
论文标题
MiniRocket: A Very Fast (Almost) Deterministic Transform for Time Series Classification
论文地址
https://dl.acm.org/doi/abs/10.1145/3447548.3467231
源码链接
https://github.com/angus924/minirocket
02
论文标题
Deep Learning Embeddings for Data Series Similarity Search
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467317
源码链接
https://github.com/qtwang/SEAnet
03
论文标题
Fast and Accurate Partial Fourier Transform for Time Series Data
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467293
源码链接
https://github.com/snudatalab/PFT
04
论文标题
A Transformer-based Framework for Multivariate Time Series Representation Learning
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467401
源码链接
https://github.com/gzerveas/mvts_transformer
05
论文标题
ST-Norm: Spatial and Temporal Normalization for Multi-variate Time Series Forecasting
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467330
源码链接
https://github.com/JLDeng/ST-Norm
06
论文标题
Statistical Models Coupling Allows for Complex Local Multivariate Time Series Analysis
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467362
源码链接
暂未开源
07
论文标题
Causal and Interpretable Rules for Time Series Analysis
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467161
源码链接
暂未开源
08
论文标题
Multivariate Time Series Anomaly Detection and Interpretation using Hierarchical Inter-Metric and Temporal Embedding
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467075
源码链接
https://github.com/zhhlee/InterFusion
09
论文标题
Practical Approach to Asynchronous Multivariate Time Series Anomaly Detection and Localization
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467174
源码链接
https://github.com/eBay/RANSynCoders
10
论文标题
Time Series Anomaly Detection for Cyber-physical Systems via Neural System Identification and Bayesian Filtering
论文地址
https://dl.acm.org/doi/10.1145/3447548.3467137
源码链接
https://github.com/NSIBF/NSIBF
IJCAI
国际人工智能联合会议 IJCAI（International Joint Conference on Artificial Intelligence）主要由国际人工智能联合会议组织和东道国国家人工智能学会联合主办（每两年举行一次），旨在通过会议记录、书籍、录像和教材的方式传播人工智能在会议上提出了尖端的科学成果。被中国计算机协会CCF推荐为A类会议。
IJCAI 2021 论文接收成果：共收到 4204 篇投稿，其中 587 篇论文被接收，接收率为 13.9%。相比于 IJCAI 2020 接受率 12.6%（592/4717）有所上升。
以下主要梳理了IJCAI 2021中Main Track & Survey Papers & Journal Papers的时间序列相关前沿研究成果。
主要方向包括：时间序列预测，时间序列分类，时序因果挖掘，多元时序分析，时序表示学习等。
01
论文标题
Time-Aware Multi-Scale RNNs for Time Series Modeling
论文地址
https://www.ijcai.org/proceedings/2021/315
源码链接
https://github.com/qianlima-lab/TAMS-RNNs
02
论文标题
Time-Series Representation Learning via Temporal and Contextual Contrasting
论文地址
https://www.ijcai.org/proceedings/2021/324
源码链接
https://github.com/emadeldeen24/TS-TCC
03
论文标题
Adversarial Spectral Kernel Matching for Unsupervised Time Series Domain Adaptation
论文地址
https://www.ijcai.org/proceedings/2021/378
源码链接
https://github.com/jarheadjoe/Adv-spec-ker-matching
04
论文标题
Two Birds with One Stone: Series Saliency for Accurate and Interpretable Multivariate Time Series Forecasting
论文地址
https://www.ijcai.org/proceedings/2021/397
源码链接
暂未开源
05
论文标题
TE-ESN: Time Encoding Echo State Network for Prediction Based on Irregularly Sampled Time Series Data
论文地址
https://www.ijcai.org/proceedings/2021/414
源码链接
暂未开源
06
论文标题
Multi-series Time-aware Sequence Partitioning for Disease Progression Modeling
论文地址
https://www.ijcai.org/proceedings/2021/493
源码链接
暂未开源
07
论文标题
Time Series Data Augmentation for Deep Learning: A Survey
论文地址
https://www.ijcai.org/proceedings/2021/631
源码链接
暂未开源
08
论文标题
Uncertain Time Series Classification
论文地址
https://www.ijcai.org/proceedings/2021/683
源码链接
https://github.com/frankl1/ustc
09
论文标题
Learning Temporal Causal Sequence Relationships from Real-Time Time-Series
论文地址
https://arxiv.org/abs/1905.12262
源码链接
暂未开源
本次推荐的论文够大家看一周了叭？本来学姐想把这几个会议拆开来发的，又怕大家觉得推的太少了，就一次性做了汇总！
以下资料：
关注公众号【学姐带你玩AI】进AI交流群&领取学习资料
①拿来就能跑通的论文复现代码讲解及数据集。
②获得视频课程的课件及思维导图
③加入up主AI学习交流群。
④能写进简历的企业级项目实战。
⑤从0到1带你一步步成为数据分析师。
⑥了解各种比赛baseline，提分的trick技巧。
⑦快速入门人工智能，夯实算法和代码基础。
⑧全部135+篇论文复现讲解视频。
⑨专属学习路径规划，学习困惑解答。
⑩不定期免费前沿论文直播。
文章内容来源网站：https://dreamhomes.top/
免责声明：所载内容来源互联网，仅供参考。转载稿件版权归原作者和机构所有，如有侵权，请联系我们删除。
