Proximity Forest

Proximity Forest is a family of similarity-based classifiers designed to address challenges in various machine learning tasks, particularly time series classification and image recognition. Current research focuses on improving the efficiency and accuracy of these classifiers through advancements in similarity measures (like Dynamic Time Warping with tunable cost functions), incorporating multiple data channels for enhanced robustness, and developing novel algorithms to mitigate biases in data augmentation techniques. These improvements have led to state-of-the-art performance in specific applications, such as time series classification and contact tracing, highlighting the importance of proximity-based methods for diverse data types.