《Learnable Dynamic Temporal Pooling for Time Series Classification》论文解读

开一个新栏，主要是记录一下本人读过的论文总结

本篇文章是这个新栏的第一篇，是AAAI2021顶级会议中一篇关于时间序列分类中卷积神经网络(CNN)池化层的一种优化方法。

Learnable Dynamic Temporal Pooling for Time Series Classification
一种用于时间序列分类的可学习动态时间池化方法

文章链接

本文摘要：

With the increase of available time series data, predicting their class labels has been one of the most important challenges in a wide range of disciplines. Recent studies on time series classification show that convolutional neural networks (CNN) achieved the state-of-the-art performance as a single classifier. In this work, pointing out that the global pooling layer that is usually adopted by existing CNN classifiers discards the temporal information of high-level features, we present a dynamic temporal pooling (DTP) technique that reduces the temporal size of hidden representations by aggregating the features at the segment-level. For the partition of a whole series into multiple segments, we utilize dynamic time warping (DTW) to align each time point in a temporal order with the prototypical features of the segments, which can be optimized simultaneously with the network parameters of CNN classifiers. The DTP layer combined with a fullyconnected layer helps to extract further discriminative features considering their temporal position within an input time series. Extensive experiments on both univariate and multivariate time series datasets show that our proposed pooling significantly improves the classification performance.

本文作者介绍了对于时间序列预测类别中使用卷积神经网络(CNN)分类器，普遍采用全局池化的方法，这种方法在从全局进行pooling操作，就会造成一定量的高级特征时间信息丢失问题。作者进而提出了一种动态时间pooling的方法(DTP)，通过对整个时间序列进行划分片段，利用动态时间规整DTW(Dynamic Time Warping)将每个时间点按时间顺序与片段的典型特征对齐，从而在分段级别上聚合特征，减少隐藏表示的时间大小。

基础知识
Differentiable Dynamic TimeWarping 动态时间规整算法
基于时间一致性的点对点匹配用于测量两个不同长度时间序列之间的距离。

举个例子：在两段音频中，音频内容是很相似的，但是音频的长度不同，当我们需要对两段音频进行相似计算的时候，就需要先把两段音频长度通过某种方法拉长到同一长度再进行比较，这样计算的误差会有效降低。

计算方法：为了解决拉长序列长度，通常使用欧式距离计算出每个序列的每两个点之间的距离。设两个序列分别为X={X1,X2,…,Xa},Y={Y1,Y2,…,Yb}.
通过公式：