Self-Supervised Random Forest on Transformed Distribution for Anomaly Detection
Jiabin Liu, Huadong Wang, Hanyuan Hang +3 more
Anomaly detection, the task of differentiating abnormal data points from normal ones, presents a significant challenge in the realm of machine learning. Numerous strategies have been proposed to tackle this task, with classification-based methods, specifically those utilizing a self-supervised approach via random affine transformations (RATs), demonstrating remarkable performance on both image and non-image data. However, these methods encounter a notable bottleneck, the overlap of constructed labeled datasets across categories, which hampers the subsequent classifiers' ability to detect anomalies. Consequently, the creation of an effective data distribution becomes the pivotal factor for success. In this article, we introduce a model called "self-supervised forest (sForest)," which leverages the random Fourier transform (RFT) and random orthogonal rotations to craft a controlled data…