scholarly journals Key kinematic features in early training predict performance of adult female mice in a single pellet reaching and grasping task

2021 ◽  
Author(s):  
Michael Mykins ◽  
Eric Wade ◽  
Xu An ◽  
Billy You Bun Lau ◽  
Keerthi Krishnan
Keyword(s):  
Skill Acquisition ◽  
Time Series Data ◽  
Food Pellet ◽  
Maximum Velocity ◽  
Series Data ◽  
Fine Motor ◽  
Learning Approaches ◽  
Fine Motor Skill ◽  
Kinematic Parameters ◽  
Single Pellet

Detailed analyses of overly trained animal models have been long employed to decipher foundational features of skilled motor tasks and their underlying neurobiology. However, initial trial-and-error features that ultimately give rise to skilled, stereotypic movements, and the underlying neurobiological basis of flexibility in learning, to stereotypic movement in adult animals are still unclear. Knowledge obtained from addressing these questions is crucial to improve quality of life in patients affected by movement disorders. We sought to determine if known kinematic parameters of skilled movement in humans could predict learning of motor efficiency in mice during the single pellet reaching and grasping assay. Mice were food restricted to increase motivation to reach for a high reward food pellet. Their attempts to retrieve the pellet were recorded for 10 minutes a day for continuous 4 days. Individual successful and failed reaches for each mouse were manually tracked using Tracker Motion Analysis Software to extract time series data and kinematic features. We found the number of peaks and time to maximum velocity were strong predictors of individual variation in failure and success, respectively. Overall, our approach validates the use of select kinematic features to describe fine motor skill acquisition in mice and establishes peaks and time to maximum velocity as predictive measure of natural variation in motion efficiency in mice. This manually curated dataset, and kinematic parameters would be useful in comparing with pose estimation generated from deep learning approaches.

scholarly journals Tourism Demand Forecasting Based on an LSTM Network and Its Variants

Algorithms ◽  
2021 ◽  
Vol 14 (8) ◽  
pp. 243
Author(s):  
Shun-Chieh Hsieh
Keyword(s):  
Adaptive Learning ◽  
Time Series Data ◽  
Short Term Memory ◽  
Demand Forecasting ◽  
Series Data ◽  
Learning Approaches ◽  
Tourism Demand ◽  
Long Short Term Memory ◽  
Lstm Network ◽  
Gated Recurrent Unit

The need for accurate tourism demand forecasting is widely recognized. The unreliability of traditional methods makes tourism demand forecasting still challenging. Using deep learning approaches, this study aims to adapt Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Gated Recurrent Unit networks (GRU), which are straightforward and efficient, to improve Taiwan’s tourism demand forecasting. The networks are able to seize the dependence of visitor arrival time series data. The Adam optimization algorithm with adaptive learning rate is used to optimize the basic setup of the models. The results show that the proposed models outperform previous studies undertaken during the Severe Acute Respiratory Syndrome (SARS) events of 2002–2003. This article also examines the effects of the current COVID-19 outbreak to tourist arrivals to Taiwan. The results show that the use of the LSTM network and its variants can perform satisfactorily for tourism demand forecasting.

scholarly journals MITRE: predicting host status from microbiota time-series data

2018 ◽  
Author(s):  
Elijah Bogart ◽  
Richard Creswell ◽  
Georg K. Gerber
Keyword(s):  
Machine Learning ◽  
Time Series ◽  
Time Series Data ◽  
Synthetic Data ◽  
Black Box ◽  
Series Data ◽  
Learning Approaches ◽  
Rule Engine ◽  
Microbiome Composition ◽  
Host Status

AbstractLongitudinal studies are crucial for discovering casual relationships between the microbiome and human disease. We present Microbiome Interpretable Temporal Rule Engine (MITRE), the first machine learning method specifically designed for predicting host status from microbiome time-series data. Our method maintains interpretability by learning predictive rules over automatically inferred time-periods and phylogenetically related microbes. We validate MITRE’s performance on semi-synthetic data, and five real datasets measuring microbiome composition over time in infant and adult cohorts. Our results demonstrate that MITRE performs on par or outperforms “black box” machine learning approaches, providing a powerful new tool enabling discovery of biologically interpretable relationships between microbiome and human host.

Fatigue Detection for Human Aware Adaptation in Human-Robot Collaboration

2021 ◽  
Author(s):  
Rakesh Suresh Kumar ◽  
Sri Sadhan Jujjavarapu ◽  
Lung Hao Lee ◽  
Ehsan T. Esfahani
Keyword(s):  
Upper Limb ◽  
Motor Coordination ◽  
Time Series Data ◽  
Industrial Robot ◽  
Computational Effort ◽  
Series Data ◽  
Fine Motor ◽  
Riemann Geometry ◽  
Human Fatigue ◽  
Human Robot Collaboration

Abstract Knowledge about human cognitive and physical state is a key factor in physical Human-robot collaboration (pHRC). Such information benefits the robot in planning an adaptive control strategy to prevent or mitigate human fatigue. In this paper, we present a method to detect upper limb muscle fatigue during pHRC using a low-cost myoelectric sensor. We used Riemann geometry to extract robust features from the time-series data and designed a classifier to detect the binary state of human fatigue i.e. fatigued vs not fatigued. We evaluated the method using a fine-motor coordination task for the human to guide an industrial robot along a virtual path for sometime followed by a muscle curl exercise until it induces fatigue in the muscles, and then repeat the robot experiment. We recruited nine participants for the study and recorded muscle activity from their dominant upper limb using the myoelectric sensor and used the data to develop a classifier. We compared the accuracy and robustness of the classifier against conventional time-domain and wavelet-based features and showed that Riemann geometry-based features yield higher classification accuracy (∼ 91%) compared to conventional features and require less computational effort. Such classifier can be used in real-time to develop a human-aware adaptation strategy to prevent fatigue.

Effects of Quiet Mind Training on Alpha Power Suppression and Fine Motor Skill Acquisition

2020 ◽  
pp. 1-10
Author(s):  
Ebrahim Norouzi ◽  
Fatemeh Sadat Hosseini ◽  
Mohammad Vaezmosavi ◽  
Markus Gerber ◽  
Uwe Pühse ◽  
...  
Keyword(s):  
Skill Acquisition ◽  
Motor Skill ◽  
Fine Motor ◽  
Alpha Power ◽  
Fine Motor Skill ◽  
Motor Skill Acquisition ◽  
Mind Training

scholarly journals CLASSIFICATION OF CROPLANDS THROUGH FUSION OF OPTICAL AND SAR TIME SERIES DATA

2016 ◽  
Vol XLI-B7 ◽  
pp. 703-704 ◽  
Author(s):  
S. Park ◽  
J. Im
Keyword(s):  
Time Series ◽  
Vegetation Index ◽  
Time Series Data ◽  
Accuracy Assessment ◽  
Sensor Data ◽  
Series Data ◽  
Support Vector ◽  
Learning Approaches ◽  
Classification Problems ◽  
Climate Conditions

Many satellite sensors including Landsat series have been extensively used for land cover classification. Studies have been conducted to mitigate classification problems associated with the use of single data (e.g., such as cloud contamination) through multi-sensor data fusion and the use of time series data. This study investigated two areas with different environment and climate conditions: one in South Korea and the other in US. Cropland classification was conducted by using multi-temporal Landsat 5, Radarsat-1 and digital elevation models (DEM) based on two machine learning approaches (i.e., random forest and support vector machines). Seven classification scenarios were examined and evaluated through accuracy assessment. Results show that SVM produced the best performance (overall accuracy of 93.87%) when using all temporal and spectral data as input variables. Normalized Difference Water Index (NDWI), SAR backscattering, and Normalized Difference Vegetation Index (NDVI) were identified as more contributing variables than the others for cropland classification.

scholarly journals Deeptime: a Python library for machine learning dynamical models from time series data

2021 ◽  
Author(s):  
Moritz Hoffmann ◽  
Martin Konrad Scherer ◽  
Tim Hempel ◽  
Andreas Mardt ◽  
Brian de Silva ◽  
...  
Keyword(s):  
Time Series ◽  
Time Series Data ◽  
Ease Of Use ◽  
Series Data ◽  
Relaxation Processes ◽  
Learning Approaches ◽  
Dynamical Models ◽  
Collective Variables ◽  
Deep Model ◽  
State Models

Abstract Generation and analysis of time-series data is relevant to many quantitative fields ranging from economics to fluid mechanics. In the physical sciences, structures such as metastable and coherent sets, slow relaxation processes, collective variables, dominant transition pathways or manifolds and channels of probability flow can be of great importance for understanding and characterizing the kinetic, thermodynamic and mechanistic properties of the system. Deeptime is a general purpose Python library offering various tools to estimate dynamical models based on time-series data including conventional linear learning methods, such as Markov state models (MSMs), Hidden Markov Models and Koopman models, as well as kernel and deep learning approaches such as VAMPnets and deep MSMs. The library is largely compatible with scikit-learn, having a range of Estimator classes for these different models, but in contrast to scikit-learn also provides deep Model classes, e.g. in the case of an MSM, which provide a multitude of analysis methods to compute interesting thermodynamic, kinetic and dynamical quantities, such as free energies, relaxation times and transition paths. The library is designed for ease of use but also easily maintainable and extensible code. In this paper we introduce the main features and structure of the deeptime software. Deeptime can be found under https://deeptime-ml.github.io/.

scholarly journals Multifocal stimulation of the cerebro-cerebellar loop during the acquisition of a novel motor skill

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Maximilian J. Wessel ◽  
Chang-hyun Park ◽  
Elena Beanato ◽  
Estelle A. Cuttaz ◽  
Jan E. Timmermann ◽  
...  
Keyword(s):  
Motor Learning ◽  
Skill Acquisition ◽  
Motor Skill ◽  
Structural Integrity ◽  
Brain Network ◽  
Fine Motor ◽  
Superior Cerebellar Peduncle ◽  
Fine Motor Skill ◽  
Double Blind ◽  
Cerebellar Stimulation

AbstractTranscranial direct current stimulation (tDCS)-based interventions for augmenting motor learning are gaining interest in systems neuroscience and clinical research. Current approaches focus largely on monofocal motorcortical stimulation. Innovative stimulation protocols, accounting for motor learning related brain network interactions also, may further enhance effect sizes. Here, we tested different stimulation approaches targeting the cerebro-cerebellar loop. Forty young, healthy participants trained a fine motor skill with concurrent tDCS in four sessions over two days, testing the following conditions: (1) monofocal motorcortical, (2) sham, (3) monofocal cerebellar, or (4) sequential multifocal motorcortico-cerebellar stimulation in a double-blind, parallel design. Skill retention was assessed after circa 10 and 20 days. Furthermore, potential underlying mechanisms were studied, applying paired-pulse transcranial magnetic stimulation and multimodal magnetic resonance imaging-based techniques. Multisession motorcortical stimulation facilitated skill acquisition, when compared with sham. The data failed to reveal beneficial effects of monofocal cerebellar or additive effects of sequential multifocal motorcortico-cerebellar stimulation. Multimodal multiple linear regression modelling identified baseline task performance and structural integrity of the bilateral superior cerebellar peduncle as the most influential predictors for training success. Multisession application of motorcortical tDCS in several daily sessions may further boost motor training efficiency. This has potential implications for future rehabilitation trials.

scholarly journals Classification of Micro-Damage in Piezoelectric Ceramics Using Machine Learning of Ultrasound Signals

Sensors ◽  
2019 ◽  
Vol 19 (19) ◽  
pp. 4216 ◽  
Author(s):  
Gaurav Tripathi ◽  
Habib Anowarul ◽  
Krishna Agarwal ◽  
Dilip Prasad
Keyword(s):  
Machine Learning ◽  
Classification Accuracy ◽  
Nearest Neighbor ◽  
Time Series Data ◽  
Series Data ◽  
Learning Approaches ◽  
K Nearest Neighbor ◽  
Power Spectral ◽  
Slope Change ◽  
Domain Information

Ultrasound based structural health monitoring of piezoelectric material is challenging if a damage changes at a microscale over time. Classifying geometrically similar damages with a difference in diameter as small as 100 μ m is difficult using conventional sensing and signal analysis approaches. Here, we use an unconventional ultrasound sensing approach that collects information of the entire bulk of the material and investigate the applicability of machine learning approaches for classifying such similar defects. Our results show that appropriate feature design combined with simple k-nearest neighbor classifier can provide up to 98% classification accuracy even though conventional features for time-series data and a variety of classifiers cannot achieve close to 70% accuracy. The newly proposed hybrid feature, which combines frequency domain information in the form of power spectral density and time domain information in the form of sign of slope change, is a suitable feature for achieving the best classification accuracy on this challenging problem.

Dexterity of the Less Affected Hand in Children With Hemiplegic Cerebral Palsy

Hand ◽  
2021 ◽  
pp. 155894472199080
Author(s):  
Matthew B. Burn ◽  
Gloria R. Gogola
Keyword(s):  
Cerebral Palsy ◽  
Birth Weight ◽  
Skill Acquisition ◽  
Fine Motor ◽  
Normative Values ◽  
Comprehensive Treatment ◽  
Fine Motor Skill ◽  
Motor Skill Acquisition ◽  
Hemiplegic Cerebral Palsy ◽  
Treatment Plans

Background: To determine if the “unaffected” hand in children with hemiplegic cerebral palsy (CP) is truly unaffected. Methods: We performed a retrospective review of manual dexterity as measured by the Functional Dexterity Test (FDT) in 66 children (39 boys, 27 girls, mean age: 11 years 4 months) with hemiplegic CP. Data were stratified by Manual Ability Classification System (MACS) level, birth weight, and gestational age at birth, and compared with previously published normative values. Results: The FDT speed of the less affected hand is significantly lower than typically developing (TD) children ( P < .001). The development of dexterity is significantly lower than TD children (0.009 vs. 0.036 pegs/s/year, P < .001), with a deficit that increases with age. MACS score, birth weight, and age at gestation are not predictors of dexterity. The dexterity of the less affected hand is poorly correlated with that of the more affected hand. Conclusions: Both dexterity and rate of fine motor skill acquisition in the less affected hand of children with hemiplegic CP is significantly less than that of TD children. The less affected hand should be evaluated and included in comprehensive treatment plans for these children.

Sign in / Sign up

Export Citation Format

Share Document