Physical rehabilitation is crucial in healthcare, facilitating recovery from injuries or illnesses and improving overall health. However, a notable global challenge stems from the shortage of professional physiotherapists, particularly acute in some developing countries, where the ratio can be as low as one physiotherapist per 100,000 individuals. To address these challenges and elevate patient care, the field of physical rehabilitation is progressively integrating Computer Vision and Human Activity Recognition (HAR) techniques. Numerous research efforts aim to explore methodologies that assist in rehabilitation exercises and evaluate patient movements, which is crucial as incorrect exercises can potentially worsen conditions. This study investigates applying various deep-learning models for classifying exercises using the benchmark KIMORE and UI-PRMD datasets. Employing Bi-LSTM, LSTM, CNN, and CNN-LSTM, alongside a Random Search for architectural design and Hyper-parameter tuning, our investigation reveals the (CNN) model as the top performer. After applying cross-validation, the technique achieves remarkable mean testing accuracy rates of 93.08% on the KIMORE dataset and 99.7% on the UI-PRMD dataset. This marks a slight improvement of 0.75% and 0.1%, respectively, compared to previous techniques. In addition, expanding beyond exercise classification, this study explores the KIMORE dataset’s utility for disease identification, where the (CNN) model consistently demonstrates an outstanding accuracy of 89.87%, indicating its promising role in both exercises and disease identification within the context of physical rehabilitation.
Feature selection is a significant issue in the machine learning process. Most datasets include features that are not needed for the problem being studied. These irrelevant features reduce both the efficiency and accuracy of the algorithm. It is possible to think about feature selection as an optimization problem. Swarm intelligence algorithms are promising techniques for solving this problem. This research paper presents a hybrid approach for tackling the problem of feature selection. A filter method (chi-square) and two wrapper swarm intelligence algorithms (grey wolf optimization (GWO) and particle swarm optimization (PSO)) are used in two different techniques to improve feature selection accuracy and system execution time. The performance of the two phases of the proposed approach is assessed using two distinct datasets. The results show that PSOGWO yields a maximum accuracy boost of 95.3%, while chi2-PSOGWO yields a maximum accuracy improvement of 95.961% for feature selection. The experimental results show that the proposed approach performs better than the compared approaches.
27 November 2024 in Computers in Biology and Medicine
Data clustering is an unsupervised learning task that has been extensively studied, given its wide applicability in various domains. Traditional algorithms often struggle to achieve a balance between exploration and exploitation, leading to sub-optimal solutions. This paper presents a novel hybrid algorithm named SA-PSO-GK++ that synergistically combines Particle Swarm Optimization (PSO), K-means++, Simulated Annealing (SA), and Gaussian Estimation of Distribution to tackle this issue effectively. The proposed SA-PSO-GK++ aims to overcome the drawbacks of existing methods by leveraging the strengths of each individual algorithm. The K-means++ initialization reduces the risk of poor initial centroids, while PSO aids in efficient search space exploration. GED provides a statistical model of the particle space, enabling the algorithm to generate new potential solutions that are statistically guided by the current best solutions. Additionally, the incorporation of Simulated Annealing allows the algorithm to escape local minima, thereby enhancing its global search capability. We evaluate the effectiveness of SA-PSO-GK++ using benchmark datasets from the UCI Machine Learning Repository, including the Iris, Breast cancer, Heart datasets and contraceptive method choice datasets. The proposed method outperforms conventional and some of the state-of-the-art hybrid clustering algorithms in terms of sum of euclidean distance, normalized index, and error rates. These advantages make SA-PSO-GK++ a compelling option for a wide range of clustering applications. The results offer promising avenues for future research in optimizing and applying this innovative clustering technique in diverse domains.
21 September 2024 in Signal, Image and Video Processing