Dynamic Time Warping Approach Research Articles

Prediction-enabled path planning for multi-ship encounters in Oslofjord

Maritime navigation in multi-ship encounters presents challenges due to dynamic perturbations and complexity, which lead to elevated safety concerns. Conventional path-planning methods exhibit several issues. This paper addresses these issues by proposing a prediction-enabled path-planning method for multi-ship encounters in the Horten and Moss area within Oslofjord, Norway, the first industrial testbed for autonomous ferries nationwide. The method leverages action classification and trajectory projection techniques to enhance path planning with a more informed decision-making process. A convolution neural network is employed for the binary classification of the own ship’s action classes, and a dynamic time-warping approach is applied to predict the future positions of the target ships based on trajectory similarity. Experiments utilizing Automatic Identification System data from 2013 to 2019 demonstrated the method’s effectiveness. Classification results show a satisfactory F1 score, and comparative evaluations with the classic A* method showcase superior performance. Real-case tests affirm the method’s ability to provide a trackable path while ensuring a safe distance between the own ship and target ships.

Inferring the diurnal variability of OH radical concentrations over the Amazon from BVOC measurements

The atmospheric oxidation of biogenic volatile organic compounds (BVOC) by OH radicals over tropical rainforests impacts local particle production and the lifetime of globally distributed chemically and radiatively active gases. For the pristine Amazon rainforest during the dry season, we empirically determined the diurnal OH radical variability at the forest-atmosphere interface region between 80 and 325 m from 07:00 to 15:00 LT using BVOC measurements. A dynamic time warping approach was applied showing that median averaged mixing times between 80 to 325 m decrease from 105 to 15 min over this time period. The inferred OH concentrations show evidence for an early morning OH peak (07:00–08:00 LT) and an OH maximum (14:00 LT) reaching 2.2 (0.2, 3.8) × 106 molecules cm−3 controlled by the coupling between BVOC emission fluxes, nocturnal NOx accumulation, convective turbulence, air chemistry and photolysis rates. The results were evaluated with a turbulence resolving transport (DALES), a regional scale (WRF-Chem) and a global (EMAC) atmospheric chemistry model.

Stock Portfolio Optimization Using Mean-Variance and Mean Absolute Deviation Model Based On K-Medoids Clustering by Dynamic Time Warping

The tendency of investors to choose investments with maximum return and minimal risk causes the need for diversification in a portfolio to form an optimal portfolio. A lot of research on stock portfolio optimization has been conducted extensively, but not many have tried to apply machine learning concepts such as clustering analysis to accelerate the establishment of a model that can have a positive effect on the time and cost efficiency of portfolio management. However, clustering is only limited to determining the optimal stock candidate, so it is necessary to add another optimization model to calculate the portfolio weight. Based on these problems, this study carried out portfolio optimization using Mean-Variance (MV) and Mean Absolute Deviation (MAD) model based on K-Medoids Clustering by Dynamic Time Warping approach using Monte Carlo-Expected Tail Loss for risk analysis. Based on the analysis results, the MAD portfolio is more optimal than the MV portfolio by the MAD portfolio consists of five stocks, namely BMRI shares with a weight of 0.06243, UNTR shares of 0.08658, BBRI shares of 0.10285, BBCA of 0.53623, and KLBF shares of 0.21191 are the best optimal portfolios. The optimal portfolio of the MAD model has a rate of return of 87.836% in May 2017 - December 2022 with a portfolio performance of 0.03704, while the resulting risk level based on Carlo-Expected Tail Loss is 2.2416%.

Assessing Driving Styles in Commercial Motor Vehicle Drivers After Take-Over Conditions in Highly Automated Vehicles

Assessing drivers’ behavior after transition from automated to manual driving (referred as to take-over condition) in highly automated vehicles (SAE Level 4) is a widely studied area. However, analyzing Commercial Motor Vehicle (CMV) drivers’ post-take-over behavior has received less attention, whereas it is forecasted that CMVs will be the first to vastly adopt highly automated vehicle technology. This study aims to analyze and compare CMV drivers’ driving styles in take-over conditions with continuous manual driving. Assessing driving style, which is a function of various variables and actions, provides a comprehensive understanding of the changes in post-take-over behavior. Hence, the driving behaviors of 45 CMV drivers are collected using a driving simulator, and we investigated whether the driving style is subject to driving mode (take-over or manual), automation duration, repeated take-overs, and driver’s factors. Here, drivers’ driving behavior is classified into three driving styles, normal, conservative, and risky by using Multivariate Dynamic Time Warping approach followed by <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula> -means clustering. Comparing driving styles in take-over and manual driving conditions showed that conservative and risky driving styles (as in more speed reduction, harder brakes, and unsafe turns) are more common in take-over conditions. Furthermore, to gain behavioral insight into the detected driving styles, Generalized Linear Models are applied to model the driving behavior indices in each driving style. Modeling results showed that long-phase automation, traffic/environmental conditions, and bad driving history deteriorate post-take-over behavior. The findings of this paper provide valuable information to automotive companies and transportation planners on the nature of take-over conditions.

Context-aware similarity measurement of lane-changing trajectories

Lane change (LC), which is an essential behavior in driving, has an important role in traffic efficiency and safety. Examining the spatial lane change behavior of drivers and determining their movement patterns through investigation of the similarities of lane change trajectories is crucial in transportation systems. The moving objectives’ trajectories are a function of their contexts. Thus, in order to consider these contexts in studies of movements, acquiring a deeper understanding of them is necessary. This paper takes advantage of internal/external contexts and spatial footprints in order to contextualize a measure for lane change trajectory similarities. By employing dynamic time warping (DTW), the similarities of multi-dimensional trajectories were estimated, and through the Next Generation Simulation (NGSIM) dataset, the dynamic time warping approach was scrutinized. The results show that contextual information can limit and increase movements and must be utilized. The vehicles’ lane change trajectories were employed in order to present three novel prediction models for trajectory-grounded position. The historical data underwent similarity evaluations for the prediction of positions. According to the results, the suggested models were applicable in precise anticipation of the positions of LC vehicles. As far as the authors are aware, this research is the first one to employ a similarity-based framework to present a thorough analysis of position simulation of LC vehicles and similarities of lane change trajectories.

Investigating similarities between Islamic and conventional banks in GCC countries: a dynamic time warping approach

PurposeSeveral studies have studied the points that distinguish Islamic banks from conventional ones. The corresponding conclusions are a bit contradictory. This paper aims to study the similarities between Islamic and conventional banks in the Gulf countries using a new approach, namely, the clustering method based on dynamic time warping (DTW) distance.Design/methodology/approachTo study the similarities between Islamic and conventional banks, in Gulf Cooperation Council (GCC) countries, this study used the DTW distance. Then, a clustering based on this distance was carried out to find out which banks are the most similar. Finally, the authors have studied the factors that explain these similarities.FindingsThis empirical study covered 44 Islamic banks and 46 conventional banks in GCC countries during 2006–2015. The results show that Islamic and conventional banks are included in the same cluster for Qatar, Bahrain and Oman. In contrast, Islamic and conventional banks do not share the same cluster for the Kingdom of Saudi Arabia, Kuwait and the United Arab Emirates. This is because of the establishment of interest rates below discount rates. In this case, banks are incentivized to take more risks to compensate for interest losses, which increases efficiency and allocates Islamic and conventional banks to different clusters. Accordingly, there is no absolute discrimination because of the initial status between Islamic and conventional banks. However, the overall banks, either Islamic or conventional, are discriminated through the distance of the banking applied interest rate and the social discount rate.Originality/valueDTW distance-based clustering is a very suitable method for emphasizing the similarities that may exist between conventional and Islamic banks. This technique has not previously been used in the literature in question.

Measuring hidden demand and price behavior from US outbound health tourism spending

The health tourism literature has covered several topics, but studies on the workings of health tourism demand and prices are still under-highlighted. This study investigates US outbound health tourism demand and price developments (real exchange rates) for 50 destinations, using outbound health tourism spending as a reference variable. The study applies the dynamic time warping approach, a sophisticated technique to assess similarities between time series, specifically, the outbound health spending with demand and prices. The results show that demand developments have similarities with US outbound health tourism spending in the short-run, with about 1.2% of all US outbound travelers being health tourists. Price developments have both long- and short-term similarities with outbound health tourism spending. The findings could help policy-makers in better managing the health tourism industry.

A dynamic time warping approach for handling class imbalanced medical datasets with missing values: A case study of protein localization site prediction

Class imbalanced medical datasets, such as cancer prediction, contain imbalanced numbers of data in different classes leading to skewed class distribution, which makes it very difficult for a classifier to distinguish between minority (i.e. cancer) and majority (i.e. non-cancer) classes. Related studies in the literature have proposed different types of solutions for the class imbalance problem including data level, algorithmic level, and cost-sensitive learning approaches. However, none of these potential solutions have considered the issue of missing attribute values residing in the class imbalanced medical datasets, especially for the minority class. Missing value imputation is commonly used for the construction of some models where statistical or machine learning techniques are used to produce estimations to replace the missing values. However, the existing imputation methods require a certain number of observed data to produce their estimations, the major challenge for them being that the amount of observed data (with no missing values) in the minority class is very limited, or that some data are not complete. In this paper, we proposed a novel approach, namely Dynamic Time Warping-based Imputation (DTWI), to handle class imbalanced datasets with missing values. Based on the similarity measurement technique of DTW, all of the data (with or without missing values) in the minority class can be used for missing value imputation. The experimental results based on 10 different class imbalanced medical datasets show that when the missing rates in the minority classes are smaller than 30%, DTWI performs similarly to the baseline K-NN imputation method and better than the mean/mode imputation and case deletion methods. When the missing rates are larger than 30%, DTWI significantly outperform the other techniques.

Dynamic time warping approach for optimized locomotor impairment detection using biomedical signal processing

The study of walking pattern also known as gait is affected by several underlying musculoskeletal and neurological factors. All humans have contrasting gait pattern however the pattern lies within a predictable range for all human with no underlying health disorders affecting gait. However, deviation from a regular distribution range can indicate underlying health conditions. The gait could be regularly and inexpensively monitored in real-time with help of smartphones embedded sensors such as tri-axial accelerometers and tri-axial gyroscopes as smartphones are widely available. The advancements in computational power and artificial intelligence led to the development of several supervised machine learning algorithms which could be used to detect certain patterns associated with different gait conditions. However, the machine learning has widely been used to study different gait diseases, with this paper we have proposed a more efficient motion tracking approach and a new algorithm utilizing Dynamic Time Warping for an optimized locomotor disorder detection with affordable, practically applicable and globally deployable setup to monitor gait. In this experiment, we have tracked motion of 30 healthy and gait deformed human subjects with the proposed motion tracking approach and utilizing the proposed algorithm, we have achieved 35.95 times faster classification with up to 95.20% detection accuracy.

Hidden Markov Model based stride segmentation on unsupervised free-living gait data in Parkinson\u2019s disease patients

BackgroundTo objectively assess a patient’s gait, a robust identification of stride borders is one of the first steps in inertial sensor-based mobile gait analysis pipelines. While many different methods for stride segmentation have been presented in the literature, an out-of-lab evaluation of respective algorithms on free-living gait is still missing.MethodTo address this issue, we present a comprehensive free-living evaluation dataset, including 146.574 semi-automatic labeled strides of 28 Parkinson’s Disease patients. This dataset was used to evaluate the segmentation performance of a new Hidden Markov Model (HMM) based stride segmentation approach compared to an available dynamic time warping (DTW) based method.ResultsThe proposed HMM achieved a mean F1-score of 92.1% and outperformed the DTW approach significantly. Further analysis revealed a dependency of segmentation performance to the number of strides within respective walking bouts. Shorter bouts (< 30 strides) resulted in worse performance, which could be related to more heterogeneous gait and an increased diversity of different stride types in short free-living walking bouts. In contrast, the HMM reached F1-scores of more than 96.2% for longer bouts (> 50 strides). Furthermore, we showed that an HMM, which was trained on at-lab data only, could be transferred to a free-living context with a negligible decrease in performance.ConclusionThe generalizability of the proposed HMM is a promising feature, as fully labeled free-living training data might not be available for many applications. To the best of our knowledge, this is the first evaluation of stride segmentation performance on a large scale free-living dataset. Our proposed HMM-based approach was able to address the increased complexity of free-living gait data, and thus will help to enable a robust assessment of stride parameters in future free-living gait analysis applications.

Multimodal Unsupervised Speech Translation for Recognizing and Evaluating Second Language Speech

This paper addresses an automatic proficiency evaluation and speech recognition for second language (L2) speech. The proposed method recognizes the speech uttered by the L2 speaker, measures a variety of fluency scores, and evaluates the proficiency of the speaker’s spoken English. Stress and rhythm scores are one of the important factors used to evaluate fluency in spoken English and are computed by comparing the stress patterns and the rhythm distributions to those of native speakers. In order to compute the stress and rhythm scores even when the phonemic sequence of the L2 speaker’s English sentence is different from the native speaker’s one, we align the phonemic sequences based on a dynamic time-warping approach. We also improve the performance of the speech recognition system for non-native speakers and compute fluency features more accurately by augmenting the non-native training dataset and training an acoustic model with the augmented dataset. In this work, we augment the non-native speech by converting some speech signal characteristics (style) while preserving its linguistic information. The proposed variational autoencoder (VAE)-based speech conversion network trains the conversion model by decomposing the spectral features of the speech into a speaker-invariant content factor and a speaker-specific style factor to estimate diverse and robust speech styles. Experimental results show that the proposed method effectively measures the fluency scores and generates diverse output signals. Also, in the proficiency evaluation and speech recognition tests, the proposed method improves the proficiency score performance and speech recognition accuracy for all proficiency areas compared to a method employing conventional acoustic models.

Dynamic Time-Warping Correction for Shifts in Ultrahigh Resolving Power Ion Mobility Spectrometry and Structures for Lossless Ion Manipulations.

Detection of arrival time shifts between ion mobility spectrometry (IMS) separations can limit achievable resolving power (Rp), particularly when multiple separations are summed or averaged, as commonly practiced in IMS. Such variations can be apparent in higher Rp measurements and are particularly evident in long path length traveling wave structures for lossless ion manipulations (SLIM) IMS due to their typically much longer separation times. Here, we explore data processing approaches employing single value alignment (SVA) and nonlinear dynamic time warping (DTW) to correct for variations between IMS separations, such as due to pressure fluctuations, to enable more effective spectrum summation for improving Rp and detection of low-intensity species. For multipass SLIM IMS separations, where narrow mobility range measurements have arrival times that can extend to several seconds, the SVA approach effectively corrected for such variations and significantly improved Rp for summed separations. However, SVA was much less effective for broad mobility range separations, such as obtained with multilevel SLIM IMS. Changes in ions' arrival times were observed to be correlated with small pressure changes, with approximately 0.6% relative arrival time shifts being common, sufficient to result in a loss of Rp for summed separations. Comparison of the approaches showed that DTW alignment performed similarly to SVA when used over a narrow mobility range but was significantly better (providing narrower peaks and higher signal intensities) for wide mobility range data. We found that the DTW approach increased Rp by as much as 115% for measurements in which 50 IMS separations over 2 s were summed. We conclude that DTW is superior to SVA for ultra-high-resolution broad mobility range SLIM IMS separations and leads to a large improvement in effective Rp, correcting for ion arrival time shifts regardless of the cause, as well as improving the detectability of low-abundance species. Our tool is publicly available for use with universal ion mobility format (.UIMF) and text (.txt) files.

Classification and Forecasting Model for Covid -19 Disease Severities based on Medical Diagnosis using Weighted Average Dynamic Time Warping Technique

Due to Covid-19 pandemic, mankind was affected with respect to mental and physical stress. The causes of the disease have to be analysed based on the correlation factors such as Medication, Age, Gender, physical fitness and habits. In this paper we have proposed a classification model based on weighted average dynamic time warping approach to detect the disease severity as High, Medium, and Low by considering the multi-variant dependent variables that affect the prediction of Covid-19 positive cases. We also proposed the forecasting model based on the time series exponential moving average to identify the growth of disease with respect to Age, Gender and Medical history of the covid-19 positive patient. The results are obtained by defining the correlation function to measure the disease severity in the range of High, Medium and Low. The time series analysis is done with respect to mean average disease severity and also number of positive cases. The forecasting is performed based on the age, gender and existing disorders in health. The results are analysed with other time series classification models such as weighted time wrapping to make the model fits maximum to the available input.

A Bayesian dynamic time warping approach for nonuniform Doppler paths in acoustic multipath channel

The estimation of Doppler effects is indispensable to many acoustic signal processing applications including reliable underwater acoustic communication. Computationally expensive ambiguity-function based methods have good performance in estimating slowly varying uniform Doppler shifts. However, for highly reverberant environments such as in shallow water acoustics, there has not been a robust approach to estimate non-uniformly time varying Doppler along each of the different propagation paths. In this paper, we propose a multidimensional dynamic time warping (DTW) algorithm to precisely estimate non-uniform time varying Doppler in acoustic multipath channels. The proposed algorithm exploits a Bayesian approach to map multidimensional temporal scaling, while guaranteeing polynomial time complexity in the length of the signal. Both simulation and experimental results for signals received from moving sources in confined reverberant spaces will be provided to show the capability of the proposed approach.

Land cover mapping based on time‐series MODIS‐NDVI using a dynamic time warping approach: A casestudy of the agricultural pastoral ecotone of northern China

AbstractTo curb land degradation, a series of ecological restoration projects have been carried out since 1999, leading to dramatic land cover change (LCC) in the agricultural pastoral ecotone of northern China (APENC). To date, there is still lack of timely and accurate land cover (LC) information for management and assessment actions. This paper presents a LC mapping scheme to map annual LC information based on dynamic time warping (DTW) approach and time‐series MODIS‐NDVI product for the APENC. The DTW approach was optimized firstly based on the change of vegetation phenology phase. Next, typical reference time‐series curves set of different LC types were established and pixel‐wise similarity was examined to identify the LC type. Before that, the reconstruction of time‐series MODIS‐NDVI was performed using harmonic analysis of time‐series and reconstruction accuracy was evaluated quantitatively at a regional scale. Results showed that the gap error was relatively large for the high vegetation cover areas while it might be ignored for the moderate and low vegetation cover areas. The overall accuracy of LC mapping in year 2010 (LC‐2010) was 77.89% and kappa coefficient reached 0.71, which suggested that LC mapping scheme could extract LC information with high accuracy. The producer's accuracy ranged between 69.83% (grasslands) and 100% (water body), while the user's accuracy changed from 58% (built‐up) to 86% (forest). The overall spatial agreement between LC‐2010 and GlobaLand30‐2010 was about 64.44%. The total area of forests, croplands, water body and built‐up exhibited an increasing trend while the contrary trend were found for the grasslands and bare land during the 2001–2017. These LC maps are valuable for research the land degradation investigation and monitoring, benefit evaluation of ecological restoration projects and effect of LCC on surface hydrothermal process in APENC.

A Data-Driven Approach for Estimating the Effects of Station Closures in Metro Systems

To estimate the impact of station closures, a novel data-driven approach is developed. First, Symbolic Aggregate Approximation (SAX) is designed to classify stations with an anomaly in passenger flow volume. Secondly, to define the trend anomaly within each segment, the Dynamic Time Warping (DTW) approach is suggested. The method simultaneously considers the mean value and trend information for passenger flow data. To check the efficiency of the proposed model, a case analysis of the Beijing metro system is illustrated. The findings suggest that the proposed model is superior in terms of the calculation of the impact of station closures to other state-of-the-art models.

Fast Constrained Dynamic Time Warping for Similarity Measure of Time Series Data

In this paper, we propose an efficient algorithm for reducing the computational complexity of dynamic time warping (DTW) for obtaining similarity measures between time series. The DTW technique exhibits superior classification accuracy compared to other algorithms but has a limitation of high computational complexity. To reduce the computational complexity of standard DTW, constrained DTW and fast DTW techniques have been proposed. The constrained DTW technique reduces the computational complexity of standard DTW by only considering limited alignments and prevent excessive alignments between two time series, which can reduce the overall classification accuracy. However, since the searching window for limited alignments is fixed, the computational complexity is still high when the length of time series is long. In contrast, fast DTW has a lower classification accuracy than the constrained DTW technique. However, fast DTW estimates the optimal alignment while considering only the alignments within an adaptive window; as two time series increase in length, the fast DTW technique more strongly reduces the computational complexity. Therefore, we propose a fast constrained DTW approach that applies the optimal alignment estimation of fast DTW within the limited alignments of constrained DTW. As the proposed fast constrained DTW operates within a fixed window area of the constrained DTW, which prevents excessive alignments, it has a classification accuracy similar to that of the constrained DTW. Also, in the fast constrained DTW, when the length of the time series is long, a low computational complexity is maintained by the influence of the adaptive window of the fast DTW. Experimental results on 19 UCR time series datasets show that the proposed fast constrained DTW method achieves computational complexity reductions of approximately 52.2% and 22.3% compared to the existing fast DTW and constrained DTW, while maintaining almost the same classification accuracy as the constrained DTW.

Iterative multiscale dynamic time warping (IMs-DTW): a tool for rainfall time series comparison

In many domains, such as weather forecasting, hydrology or civil protection, it is an important issue to characterize rainfall variability and intermittency in, either or both, a given time period or area. A variety of sensors, for instance, rain gauges, weather radars and satellites, are widely used for this purpose. Techniques to establish the similarity between rainfall time series are commonly based on the comparison of some extracted characteristic parameters (cumulative rainfall height, extreme values, rain occurrence, mean rain rate, etc.). The present study focuses on the development of a tool allowing to compare directly rainfall time series at a fine temporal scale. It allows quantifying the dissimilarity between the time series and determining a nonlinear relationship between their time axes. This study presents an algorithm based on a multiscale dynamic time warping approach, and it is based on the DTW algorithm applied on an iterative multiscale framework called IMs-DTW. This proposed algorithm is well suited for rain time series allowing point-to-point pairing between pairs of rainfall time. It takes the intermittency and the non-stationarity of the precipitation process into account. An application to measurements observed by four pluviometers located in the Paris area makes it possible to interpret the obtained results and to compare the IMs-DTW with more usual statistical features.

Analyzing precious metals returns using a Kalman smoother approach

PurposeThis paper aims to analyze the sensitivity of different external factors to the returns of the precious metals of gold, silver, platinum and palladium. The goal is to find similarities and differences between the dependencies of every factor to each metal in a time-varying framework.Design/methodology/approachA brief co-integration test for the precious metals is conducted followed by a Kalman smoother approach to study the different sensitivities to the price changes of precious metals. A dynamic time warping (DTW) approach finally compares sensitivities for pairs of precious metals to a specific factor.FindingsResults point to strong time-dependencies of the sensitivities, such as a declining relationship of gold to equity volatility. Consistent strong relationships are rare and can be identified for the consumer price index and the dollar. the DTW approach finds higher similarities between platinum and palladium compared to other pairs.Practical implicationsThe similarities and differences of the precious metals can be used by investors and risk managers in portfolio construction processes and risk analyses.Originality/valueThe focus of the research is put on a broader context of precious metals with different external factors instead of focusing on a single factor, enabling a comparison of differences and similarities of the sensitivities. The analysis via a Kalman Rauch–Tung–Striebel smoother together with a DTW approach has not been conducted before in this way and is able to characterize the dependencies by a single number.

Structured dynamic time warping for continuous hand trajectory gesture recognition

Continuous hand gesture recognition is an important area of HCI and challenged by various writing habits and unconstrained hand movement. In this paper, we propose a Structured Dynamic Time Warping (SDTW) approach for continuous hand trajectory recognition. We first propose an automatic continuous trajectory segmentation approach which combines templates and velocity information to spot the beginning and ending points in hand gesture trajectories. Then we assign different weights to feature sequences based on the structured information, from the positions of corner points in the arbitrary trajectories. Finally, we evaluate the SDTW on the Continuous Letter Trajectory (CLT) database. Experimental results show that the proposed approach is robust to the diversity of same handwritten letter, and significantly outperforms state-of-the-art approaches.