相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Computer Science, Artificial Intelligence
Sen Qiu et al.
Summary: This paper introduces common wearable sensors, smart wearable devices, and key application areas, proposing fusion methods for multi-modality and multi-location sensors. It comprehensively surveys important aspects of wearable sensor fusion methods in human activity recognition, including new technologies in unsupervised learning and transfer learning, while also discussing open research issues that need further investigation and improvement.
INFORMATION FUSION
(2022)
Article
Computer Science, Information Systems
Zhenxiang Feng et al.
Summary: This study proposes a multimode mechanism-guided product quality variable estimation model. Representative features are extracted through feature engineering and deep learning, and a multi-mode LSTM network is trained to increase adaptability. The LSTM units are cascaded to handle the multi-rate problem, and mechanism models guide the learning process. The estimation model utilizes production data, mechanism knowledge, and working condition information for improved interpretability and adaptability.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Guangyin Jin et al.
Summary: Spatial-temporal event prediction is crucial for future smart cities construction, with recent deep learning models commonly using video-like spatial-temporal modelling. However, limitations exist in considering latent region-wise correlations, prompting the proposal of the Adaptive Dual-View WaveNet (ADVW-Net) framework to capture both geographic correlations and region-wise dependencies. The integration of Convolutional Neural Network (CNN) and adaptive Graph Convolutional Neural Network (GCN) representations, along with the WaveNet architecture, allows for effective learning of long-range temporal dependencies in urban spatial-temporal event prediction.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
M. A. Castan-Lascorz et al.
Summary: This research proposes a new algorithm based on a combination of clustering, classification, and forecasting techniques for predicting both univariate and multivariate time series. The algorithm groups similar patterns of time series values using clustering and builds specific forecasting models for each pattern. Experimental results show that the proposed algorithm outperforms classical and recent forecasting methods in terms of prediction accuracy.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Zhong Zheng et al.
Summary: This paper proposes a denoising temporal convolutional recurrent autoencoder (DTCRAE) to enhance the performance of the temporal convolutional network (TCN) in time series classification (TSC). The results of computational studies demonstrate that DTCRAEs outperform other algorithms on three datasets, achieving higher accuracies and providing a better initial structure for TCN classifiers.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Ye Zhang et al.
Summary: This paper proposes a method called MSRP-IFCN for encoding time series as images for classification. The method uses multi-scale signed RP and Inception fully convolutional network to handle the scale and length variability of sequences and enhance multi-scale feature extraction. Experimental results demonstrate the superior performance of this method.
PATTERN RECOGNITION
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Ze Liu et al.
Summary: This paper introduces a Transformer architecture with a bias towards locality in video recognition, achieving a better balance between speed and accuracy compared to global self-attention mechanisms; by adapting the Swin Transformer and leveraging pre-trained models, it achieves state-of-the-art accuracy on various video recognition benchmarks.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
(2022)
Review
Computer Science, Artificial Intelligence
Alejandro Pasos Ruiz et al.
Summary: Time Series Classification (TSC) has seen significant advancements in recent years, with more focus on univariate TSC compared to multivariate TSC. While there are valuable resources for univariate TSC, there is a lack of standard test problems for MTSC. Recent bespoke MTSC algorithms, including deep learning, shapelets and bag of words approaches, have shown promise in improving classification accuracy.
DATA MINING AND KNOWLEDGE DISCOVERY
(2021)
Article
Computer Science, Artificial Intelligence
Matthew Middlehurst et al.
Summary: HIVE-COTE is a heterogeneous meta ensemble algorithm for time series classification, which has been state of the art for accuracy on the UCR time series classification archive since it was first proposed in 2016. Recently upgraded to HIVE-COTE 2.0, it introduces two new classifiers and a rocket classifier ensemble, demonstrating significantly higher accuracy than the current state of the art on multiple datasets.
Article
Computer Science, Artificial Intelligence
Zhi Chen et al.
Summary: In this paper, a time-frequency deep metric learning (TFDM) approach is proposed for Multivariate Time Series (MTS) classification. The method utilizes multilevel discrete wavelet decomposition and deep convolutional neural networks to learn time-frequency representations and incorporates a consistency regularization term to capture correlations among different levels of MTS. Experimental results demonstrate the effectiveness of the approach.
Article
Computer Science, Artificial Intelligence
Ioan-Daniel Borlea et al.
Summary: This paper introduces the novel Unified Form (UF) clustering algorithm and the Partitional Implementation of Unified Form (PIUF) algorithm, aiming to address the challenges of processing large datasets and sequential data processing. These algorithms are implemented and validated in the BigTim platform and can be applied to other data processing platforms.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Information Systems
Heshan Wang et al.
Summary: An innovative input weight establishment framework based on autoencoder theory for Echo State Networks (ESN) in time series classification tasks was proposed in this study. The Global Reversible Autoencoder (GRAE) algorithm efficiently reestablished the random initialization input weights of ESN, leading to improved performance in feature learning and classification tasks.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Ling Chen et al.
Summary: This paper introduces a deep multi-task representation learning method for time series classification and retrieval, utilizing both supervised and unsupervised information. By sharing networks, the classification and retrieval tasks can benefit from each other, leading to improved accuracy and convergence speed.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Zhiwen Xiao et al.
Summary: This paper introduces a novel multi-process collaborative architecture for time series classification, which combines multi-head convolutional neural networks and capsule mechanism to achieve better feature extraction and representation learning without the need for reconstruction to improve model accuracy.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Mehak Khan et al.
Summary: This study investigates the time series classification problem using deep learning techniques, proposing two end-to-end deep learning architectures and validating them on multiple datasets. The results show that the proposed models perform well in terms of classification testing error and f1-score, outperforming other methods and being suitable for deployment on real-time systems.
NEURAL PROCESSING LETTERS
(2021)
Article
Computer Science, Information Systems
Zhiwen Xiao et al.
Summary: Time series data contains both local and global patterns, but existing feature networks focus on local features and neglect the relationships among them. Therefore, a novel RTFN method is proposed for feature extraction in time series, consisting of TFN and LSTMaN. Experimental results show that the RTFN-based structures achieve excellent performance on multiple datasets.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Zhiyu Liang et al.
Summary: Time series classification is a critical problem in data mining with applications in various fields, leading to the proposal of numerous algorithms. Shapelet-based approaches have gained attention for their interpretability in classification results. These methods discover discriminatory subsequences and use similarity between time series and shapelets as features for building classifiers, enabling direct classification decisions based on the presence or absence of specific shapelets.
INFORMATION SCIENCES
(2021)
Proceedings Paper
Computer Science, Information Systems
Sharath M. Shankaranarayana et al.
Summary: This work introduces a novel end-to-end architecture for learning and processing tabular time-series data for tasks like time-series classification. The framework incorporates a masking technique suitable for pretraining time-series data, combines one-dimensional convolutions and transformers, and introduces timestamp embedding to handle periodic cycles and aperiodic trends.
21ST IEEE INTERNATIONAL CONFERENCE ON DATA MINING WORKSHOPS ICDMW 2021
(2021)
Article
Computer Science, Information Systems
Zechao Li et al.
Summary: This study introduces a new semi-supervised feature selection method that can select the most suitable features for each class to improve the accuracy and efficiency of data classification.
SCIENCE CHINA-INFORMATION SCIENCES
(2021)
Article
Acoustics
Zheng Lian et al.
Summary: The study proposes a multimodal learning framework for conversational emotion recognition, named conversational transformer network (CTNet). By modeling intra-modal and cross-modal interactions, capturing temporal information using lexical and acoustic features, and utilizing a bi-directional GRU component and speaker embeddings to model context-sensitive and speaker-sensitive dependencies, the experimental results demonstrate the effectiveness of the method. The approach shows a performance improvement of 2.1% to 6.2% on weighted average F1 over state-of-the-art strategies.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2021)
Article
Computer Science, Artificial Intelligence
Hassan Ismail Fawaz et al.
DATA MINING AND KNOWLEDGE DISCOVERY
(2020)
Article
Geography, Physical
Marc Russwurm et al.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2020)
Article
Computer Science, Artificial Intelligence
Fazle Karim et al.
Article
Computer Science, Artificial Intelligence
Jiancheng Sun et al.
PATTERN RECOGNITION
(2019)
Article
Engineering, Mechanical
Adriana Albu et al.
FACTA UNIVERSITATIS-SERIES MECHANICAL ENGINEERING
(2019)
Article
Computer Science, Information Systems
Jason Lines et al.
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
(2018)
Article
Computer Science, Artificial Intelligence
Isak Karlsson et al.
DATA MINING AND KNOWLEDGE DISCOVERY
(2016)
Article
Computer Science, Artificial Intelligence
Patrick Schaefer
DATA MINING AND KNOWLEDGE DISCOVERY
(2015)
Article
Computer Science, Information Systems
Houtao Deng et al.
INFORMATION SCIENCES
(2013)