Clustering Trajectories For Rats Movements Using Unsupervised Learning

The purpose of this paper is to investigate cluster trajectories for rats’ movements to monitor behavior rats before and after taking medicine to see any change in his behavior. Traditional methods take a long time and suffer from inaccuracy, such as: direct observation and video recording, which These methods are labor-intensive and vary based on the observer’s skills. There are three main stages: computer vision, image preprocessing, and unsupervised learning. The rats are placed in an open fields experiment or water maze and are detection and tracking rats movements using computer vision and cluster trajectories of rats using unsupervised learning which provides cognitive mapping via video analysis to collect insights. The YOLOv8 model was trained and tested in data sets collected from laboratory experiments in rats, achieving impressive accuracies of 92%. There are different clusters techniques and distance measures used such as: K-means in clusters and Dynamic Time Warping (DTW), Euclidean distances and P dollar Recognizer in distance measures and similarity to identify distinct movement patterns.