Time Series Analysis Using Geometric Template Matching

Multiview Unsupervised Shapelet Learning for Multivariate Time Series Clustering

Multivariate time series clustering has become an important research topic in the time series learning task, which aims to discover the correlation among multiple sequences and partition multivariate time series data into several subsets. Although there are currently some methods that can handle this task, most of them fail to discover informative subsequences from multivariate time series instances. In this paper, we first propose a novel unsupervised shapelet learning with adaptive neighbors (USLA) model for learning salient multivariate subsequences (i.e., multivariate shapelets), where the importance of each variate can be auto-determined when given a candidate multivariate shapelet. USLA performs multivariate shapelet-transformed representation learning and local structure learning simultaneously, but the performance of USLA with multivariate shapelets of different lengths is comparable to that of isometric multivariate shapelets. In fact, the shapelet-transformed representations learned from multivariate shapelets of different lengths can all represent multivariate time series instances separately and often contain complementary information to each other. Therefore, we develop a novel multiview USLA (MUSLA) model which treats shapelet-transformed representations learned from shapelets of different lengths as different views. In this way, MUSLA learns the importance of each view and the neighbor graph matrix among multiview representations when candidate multivariate shapelets of different lengths are determined. Experimental results show that MUSLA outperforms other state-of-the-art multivariate time series algorithms on real-world multivariate time series datasets.

Explainable gait recognition with prototyping encoder–decoder

Human gait is a unique behavioral characteristic that can be used to recognize individuals. Collecting gait information widely by the means of wearable devices and recognizing people by the data has become a topic of research. While most prior studies collected gait information using inertial measurement units, we gather the data from 40 people using insoles, including pressure sensors, and precisely identify the gait phases from the long time series using the pressure data. In terms of recognizing people, there have been a few recent studies on neural network-based approaches for solving the open set gait recognition problem using wearable devices. Typically, these approaches determine decision boundaries in the latent space with a limited number of samples. Motivated by the fact that such methods are sensitive to the values of hyper-parameters, as our first contribution, we propose a new network model that is less sensitive to changes in the values using a new prototyping encoder–decoder network architecture. As our second contribution, to overcome the inherent limitations due to the lack of transparency and interpretability of neural networks, we propose a new module that enables us to analyze which part of the input is relevant to the overall recognition performance using explainable tools such as sensitivity analysis (SA) and layer-wise relevance propagation (LRP).

Stamping Monitoring by Using an Adaptive 1D Convolutional Neural Network

Stamping is one of the most widely used processes in the sheet metalworking industry. Because of the increasing demand for a faster process, ensuring that the stamping process is conducted without compromising quality is crucial. The tool used in the stamping process is crucial to the efficiency of the process; therefore, effective monitoring of the tool health condition is essential for detecting stamping defects. In this study, vibration measurement was used to monitor the stamping process and tool health. A system was developed for capturing signals in the stamping process, and each stamping cycle was selected through template matching. A one-dimensional (1D) convolutional neural network (CNN) was developed to classify the tool wear condition. The results revealed that the 1D CNN architecture a yielded a high accuracy (>99%) and fast adaptability among different models.

A database of human gait performance on irregular and uneven surfaces collected by wearable sensors

Gait analysis has traditionally relied on laborious and lab-based methods. Data from wearable sensors, such as Inertial Measurement Units (IMU), can be analyzed with machine learning to perform gait analysis in real-world environments. This database provides data from thirty participants (fifteen males and fifteen females, 23.5 ± 4.2 years, 169.3 ± 21.5 cm, 70.9 ± 13.9 kg) who wore six IMUs while walking on nine outdoor surfaces with self-selected speed (16.4 ± 4.2 seconds per trial). This is the first publicly available database focused on capturing gait patterns of typical real-world environments, such as grade (up-, down-, and cross-slopes), regularity (paved, uneven stone, grass), and stair negotiation (up and down). As such, the database contains data with only subtle differences between conditions, allowing for the development of robust analysis techniques capable of detecting small, but significant changes in gait mechanics. With analysis code provided, we anticipate that this database will provide a foundation for research that explores machine learning applications for mobile sensing and real-time recognition of subtle gait adaptations.

A New Timing Error Cost Function for Binary Time Series Prediction

The ability to make predictions is central to the artificial intelligence problem. While machine learning algorithms have difficulty in learning to predict events with hundreds of time-step dependencies, animals can learn event timing within tens of trials across a broad spectrum of time scales. This suggests strongly a need for new perspectives on the forecasting problem. This paper focuses on binary time series that can be predicted within some temporal precision. We demonstrate that the sum of squared errors (SSE) calculated at every time step is not appropriate for this problem. Next, we look at the advantages and shortcomings of using a dynamic time warping (DTW) cost function. Then, we propose the squared timing error (STE) that uses DTW on the event space and applies SSE on the timing error instead of at each time step. We evaluate all three cost functions on different types of timing errors, such as phase shift, warping, and missing events, on synthetic and real-world binary time series (heartbeats, finance, and music). The results show that STE provides more information about timing error, is differentiable, and can be computed online efficiently. Finally, we devise a gradient descent algorithm for STE on a simplified recurrent neural network. We then compare the performance of the STE-based algorithm to SSE- and logit-based gradient descent algorithms on the same network architecture. The results in real-world binary time series show that the STE algorithm generally outperforms all the other cost functions considered.

Salient Subsequence Learning for Time Series Clustering

Time series has been a popular research topic over the past decade. Salient subsequences of time series that can benefit the learning task, e.g., classification or clustering, are called shapelets. Shapelet-based time series learning extracts these types of salient subsequences with highly informative features from a time series. Most existing methods for shapelet discovery must scan a large pool of candidate subsequences, which is a time-consuming process. A recent work, [1] , uses regression learning to discover shapelets in a time series; however, it only considers learning shapelets from labeled time series data. This paper proposes an Unsupervised Salient Subsequence Learning (USSL) model that discovers shapelets without the effort of labeling. We developed this new learning function by integrating the strengths of shapelet learning, shapelet regularization, spectral analysis and pseudo-label to simultaneously and automatically learn shapelets to help clustering unlabeled time series better. The optimization model is iteratively solved via a coordinate descent algorithm. Experiments show that our USSL can learn meaningful shapelets, with promising results on real-world and synthetic data that surpass current state-of-the-art unsupervised time series learning methods.

The Discriminative Power of Shape an Empirical Study in Time Series Matching

Shape provides significant discriminating power in time series matching of visual or geometric data as required in many important applications in graphics and vision. The well established dynamic time warping (DTW) algorithm and its variants do this matching by determining a non-linear time mapping to minimise euclidean distances between corresponding time-warped points. However the shape of curves is not considered. In this paper, we present a new shape-aware algorithm which uses time and shape correspondence (TSC) at increasing levels of detail to define a similarity measure with an norm to aggregate the results, making it robust to noise and missing data. The norm is implicitly regularised using a shape-based error. Through extensive experiments we empirically show that our algorithm outperforms existing state of the art algorithms, works more effectively with high dimensional data, and handles noise and missing data better. We demonstrate its versatile applicability and comparative performance using a large in-house created gait data base, an action data base from Microsoft, exercise action data from a local company, a large public time series data base from University of California, Riverside and hand movement in quaternion stream data format.

Discriminative Dimensionality Reduction for Multi-Dimensional Sequences

Since the observables at particular time instants in a temporal sequence exhibit dependencies, they are not independent samples. Thus, it is not plausible to apply i.i.d. assumption-based dimensionality reduction methods to sequence data. This paper presents a novel supervised dimensionality reduction approach for sequence data, called Linear Sequence Discriminant Analysis (LSDA). It learns a linear discriminative projection of the feature vectors in sequences to a lower-dimensional subspace by maximizing the separability of the sequence classes such that the entire sequences are holistically discriminated. The sequence class separability is constructed based on the sequence statistics, and the use of different statistics produces different LSDA methods. This paper presents and compares two novel LSDA methods, namely M-LSDA and D-LSDA. M-LSDA extracts model-based statistics by exploiting the dynamical structure of the sequence classes, and D-LSDA extracts the distance-based statistics by computing the pairwise similarity of samples from the same sequence class. Extensive experiments on several different tasks have demonstrated the effectiveness and the general applicability of the proposed methods.

Wearable Device-Based Gait Recognition Using Angle Embedded Gait Dynamic Images and a Convolutional Neural Network

The widespread installation of inertial sensors in smartphones and other wearable devices provides a valuable opportunity to identify people by analyzing their gait patterns, for either cooperative or non-cooperative circumstances. However, it is still a challenging task to reliably extract discriminative features for gait recognition with noisy and complex data sequences collected from casually worn wearable devices like smartphones. To cope with this problem, we propose a novel image-based gait recognition approach using the Convolutional Neural Network (CNN) without the need to manually extract discriminative features. The CNN’s input image, which is encoded straightforwardly from the inertial sensor data sequences, is called Angle Embedded Gait Dynamic Image (AE-GDI). AE-GDI is a new two-dimensional representation of gait dynamics, which is invariant to rotation and translation. The performance of the proposed approach in gait authentication and gait labeling is evaluated using two datasets: (1) the McGill University dataset, which is collected under realistic conditions; and (2) the Osaka University dataset with the largest number of subjects. Experimental results show that the proposed approach achieves competitive recognition accuracy over existing approaches and provides an effective parametric solution for identification among a large number of subjects by gait patterns.

Inertial Sensor-Based Gait Recognition: A Review

With the recent development of microelectromechanical systems (MEMS), inertial sensors have become widely used in the research of wearable gait analysis due to several factors, such as being easy-to-use and low-cost. Considering the fact that each individual has a unique way of walking, inertial sensors can be applied to the problem of gait recognition where assessed gait can be interpreted as a biometric trait. Thus, inertial sensor-based gait recognition has a great potential to play an important role in many security-related applications. Since inertial sensors are included in smart devices that are nowadays present at every step, inertial sensor-based gait recognition has become very attractive and emerging field of research that has provided many interesting discoveries recently. This paper provides a thorough and systematic review of current state-of-the-art in this field of research. Review procedure has revealed that the latest advanced inertial sensor-based gait recognition approaches are able to sufficiently recognise the users when relying on inertial data obtained during gait by single commercially available smart device in controlled circumstances, including fixed placement and small variations in gait. Furthermore, these approaches have also revealed considerable breakthrough by realistic use in uncontrolled circumstances, showing great potential for their further development and wide applicability.