TOD: GPU-accelerated Outlier Detection via Tensor Operations

We propose TOD, a system for efficient and scalable outlier detection (OD) on distributed multi-GPU machines. A key idea behind TOD is decomposing OD applications into basic tensor algebra operations. This decomposition enables TOD to accelerate OD computations by leveraging recent advances in deep learning infrastructure in both hardware and software. Moreover, to deploy costly OD algorithms on modern GPUs with limited memory, we introduce two key techniques. First, provable quantization speeds up OD computation and reduces its memory footprint by performing specific floating-point operations in lower precision while provably guaranteeing no accuracy loss. Second, to exploit the aggregated compute resources and memory capacity of multiple GPUs, we introduce automatic batching, which decomposes OD computations into small batches for parallel execution on multiple GPUs. TOD supports a comprehensive and diverse set of OD algorithms, e.g., LOF, PCA, and HBOS, and utility functions. Extensive evaluation on both real and synthetic OD datasets shows that TOD is on average 11.6x faster than the leading CPU-based OD system PyOD (with a maximum speedup of 38.9x), and can handle much larger datasets than various GPU baselines. Notably, TOD allows straightforward integration of additional OD algorithms and provides a unified framework for combining classical OD algorithms with deep learning methods. These combinations result in an infinite number of OD methods, many of which are novel and can be easily prototyped in TOD.