This research identifies a gap in weakly-labelled multivariate time-series classification (TSC), where state-of-the-art TSC models do not per-form well. Weakly labelled time-series are time-series containing noise and significant redundancies. In response to this gap, this paper proposes an approach of exploiting context relevance of subsequences from previous subsequences to improve classification accuracy. To achieve this, state-of-the-art Attention algorithms are experimented in combination with the top CNN models for TSC (FCN and ResNet), in an CNN-LSTM architecture. Attention is a popular strategy for context extraction with exceptional performance in modern sequence-to-sequence tasks. This paper shows how attention algorithms can be used for improved weakly labelledTSC by evaluating models on a multivariate EEG time-series dataset obtained using a commercial Emotiv headsets from participants performing various activities while driving. These time-series are segmented into sub-sequences and labelled to allow supervised TSC.
Multi-Task Learning for Interpretable Weakly Labelled Sound Event Detection
Carnegie Mellon University（卡内基梅隆大学）
近年来，弱标记学习由于其扩展声音事件检测（SED）的潜力而引起了广泛的关注，并被公式化为多实例学习（MIL）问题。本文提出了一个多任务学习（MTL）框架，从弱标记音频数据，其中包括传统的MIL设置学习。为了展示所提出的框架的效用，我们使用输入时频表示（T-F）重建作为辅助任务。我们发现，所选择的辅助任务去噪内部T-F表示，并提高了SED性能在嘈杂的录音。我们的第二个贡献是引入两步注意力池机制。通过在注意机制中具有2步，网络保留更好的T-F电平信息而不损害SED性能。第一步和第二步注意力权重的可视化有助于在T-F域中定位音频事件。为了评估所提出的框架，我们将DCASE 2019任务1声学场景数据与DCASE 2018任务2在0，10和20 db SNR下的声音事件数据进行了混合，从而产生了多类弱标记SED问题。所提出的总体框架在所有SNR上都优于现有的基准模型，特别是在0、10和20 dB SNR上分别比基准模型提高22.3%、12.8%和5.9%。我们进行了消融研究，以确定每个辅助任务和两步注意力池的SED性能改善的贡献。代码公开发布
MarketGANs: Multivariate financial time-series data augmentation using generative adversarial networks
Department of Mathematics, Sungkyunkwan University, Republic of Korea（淑阳大学数学系）；Department of Financial Engineering, Ajou University, Republic of Korea（东亚大学金融工程系）；Department of FinTech, SKK Business School, Sungkyunkwan University, Republic of Korea（淑阳大学SKK商学院金融科技系）
