A kernel for time series based on global alignments

Abstract

We propose in this paper a new family of kernels to handle times series, notably speech data, within the framework of kernel methods which includes popular algorithms such as the Support Vector Machine. These kernels elaborate on the well known Dynamic Time Warping (DTW) family of distances by considering the same set of elementary operations, namely substitutions and repetitions of tokens, to map a sequence onto another. Associating to each of these operations a given score, DTW algorithms use dynamic programming techniques to compute an optimal sequence of operations with high overall score. In this paper we consider instead the score spanned by all possible alignments, take a smoothed version of their maximum and derive a kernel out of this formulation. We prove that this kernel is positive definite under favorable conditions and show how it can be tuned effectively for practical applications as we report encouraging results on a speech recognition task.

Cited by in corpus (34)

• Learning Factorized Multimodal Representations
• Intelligent Traffic Monitoring Systems for Vehicle Classification: A Survey
• Soft-DTW: a Differentiable Loss Function for Time-Series
• On Recursive Edit Distance Kernels with Application to Time Series Classification
• Emotional Expression Classification using Time-Series Kernels
• Metric Learning for Temporal Sequence Alignment
• Autoregressive Kernels For Time Series
• Differentiable Divergences Between Time Series
• Probabilistic Time Series Forecasting with Structured Shape and Temporal Diversity
• Training Robust Deep Models for Time-Series Domain: Novel Algorithms and Theoretical Analysis
• Positive Definite Kernels in Machine Learning
• Times series averaging and denoising from a probabilistic perspective on time-elastic kernels
• Times series averaging from a probabilistic interpretation of time-elastic kernel
• Time Series Alignment with Global Invariances
• Seglearn: A Python Package for Learning Sequences and Time Series
• Kernels for sequentially ordered data
• Distribution Regression for Sequential Data
• Human Motion Analysis with Deep Metric Learning
• An Unsupervised Multivariate Time Series Kernel Approach for Identifying Patients with Surgical Site Infection from Blood Samples
• Learnable Structural Semantic Readout for Graph Classification
• On a Family of Decomposable Kernels on Sequences
• Kernels for time series with irregularly-spaced multivariate observations
• Learnable Dynamic Temporal Pooling for Time Series Classification
• The Directional Bias Helps Stochastic Gradient Descent to Generalize in Kernel Regression Models
• A Global Alignment Kernel based Approach for Group-level Happiness Intensity Estimation
• Space-efficient Feature Maps for String Alignment Kernels
• Ensemble of Hankel Matrices for Face Emotion Recognition
• Positivity and Transportation
• Efficient Kernel-based Subsequence Search for User Identification from Walking Activity
• On the separation of shape and temporal patterns in time series -Application to signature authentication-
• Interpretable Multiple-Kernel Prototype Learning for Discriminative Representation and Feature Selection
• Adaptive Down-Sampling and Dimension Reduction in Time Elastic Kernel Machines for Efficient Recognition of Isolated Gestures
• MLAT: Metric Learning for kNN in Streaming Time Series
• Graph kernels between point clouds