Flores, Fritz Kevin.
Gait abnormality detection using multimodal sensors and machine learning / Fritz Kevin Flores [and four others].
Gait can be defined as how a person walks. When the person is not able to walk properly due to different factors it can be deduced that their gait is abnormal. The proponents want to simplify and improve the process of detecting whether a person has gait abnormality or not through the use of technologies. In this study, a Kinect sensor and a smart flooring sensor were used together to collect the data needed for the study. Overall, the researchers had 30 samples containing 20 participants with an additional of 10 acted data. Features such as Stride length, Symmetry, and Cadence were collected. Then, the data were preprocessed by using methods such as feature scaling, feature extraction, and feature selection before being fed into K-Nearest Neighbors (KNN) and K-means Clustering machine learning algorithms, as well as Convolutional Neural Network (CNN) deep learning algorithms, to create different classifiers for gait. Finally, evaluation metrics such as accuracy, DBI, and loss were used toselect the best classifiers that would be used in the decision-level fusion.
Gait disorders.
Machine learning.
Detectors.
Gait abnormality detection using multimodal sensors and machine learning / Fritz Kevin Flores [and four others].
Gait can be defined as how a person walks. When the person is not able to walk properly due to different factors it can be deduced that their gait is abnormal. The proponents want to simplify and improve the process of detecting whether a person has gait abnormality or not through the use of technologies. In this study, a Kinect sensor and a smart flooring sensor were used together to collect the data needed for the study. Overall, the researchers had 30 samples containing 20 participants with an additional of 10 acted data. Features such as Stride length, Symmetry, and Cadence were collected. Then, the data were preprocessed by using methods such as feature scaling, feature extraction, and feature selection before being fed into K-Nearest Neighbors (KNN) and K-means Clustering machine learning algorithms, as well as Convolutional Neural Network (CNN) deep learning algorithms, to create different classifiers for gait. Finally, evaluation metrics such as accuracy, DBI, and loss were used toselect the best classifiers that would be used in the decision-level fusion.
Gait disorders.
Machine learning.
Detectors.