Overview of Data Augmentation Techniques in Time Series Analysis
This paper explores the transformative impact of data augmentation on time series analysis, highlighting classic methods such as interpolation and advanced techniques like GANs and VAEs. It examines the challenges, including computational costs and risks of overfitting, while showcasing the advantages of these methods. Offering a balanced perspective, the paper serves as a comprehensive guide for researchers and practitioners to address challenges and drive advancements in machine learning and deep learning applications for time series data.
1 janv. 2024
Spatiotemporal Clustering of Human Locomotion Neuropsychological Assessment in Virtual Reality Using Multi-step Model
This paper introduces a spatiotemporal clustering model to analyze neuropsychological assessments conducted in a virtual reality environment. The study focuses on clustering human navigation behaviors using a deep learning framework that separates feature extraction and clustering stages. By applying a 1D-DCAE autoencoder for feature extraction and a soft temporal clustering mechanism, the method identifies five distinct clusters corresponding to clinical groups. This research contributes to understanding navigation patterns in cognitive impairments and advances VR-based neuropsychological assessments.
1 janv. 2023
Residual Neural Network Architecture for Identifying Vestibular Disease Based on Head Kinematic Characteristics (Velocity)
1 janv. 2023
Joint Unsupervised Deep Temporal Clustering for Modeling Human Behavior in Vestibular Dysfunction: A Study of Navigation Pattern
1 janv. 2023
Evaluating the Efficiency of Multilayer Perceptron Neural Network Architecture in Classifying Cognitive Impairments Related to Human Bipedal Spatial Navigation
1 janv. 2023
Computational Analysis of Human Navigation Trajectories in the VR Magic Carpet ™ Using K-Means
1 janv. 2023