MSS-PAE: Saving Autoencoder-based Outlier Detection from Unexpected Reconstruction
This work addresses limitations in autoencoder-based outlier detection for applications requiring robust anomaly identification, representing an incremental improvement with specific gains.
The paper tackled the problem of overconfident decisions and unexpected reconstruction results in autoencoder-based outlier detection by proposing Weighted Negative Logarithmic Likelihood (WNLL) and Mean-Shift Scoring (MSS) methods, achieving a 41% relative performance improvement over the best baseline on 32 real-world datasets.
AutoEncoders (AEs) are commonly used for machine learning tasks due to their intrinsic learning ability. This unique characteristic can be capitalized for Outlier Detection (OD). However conventional AE-based methods face the issue of overconfident decisions and unexpected reconstruction results of outliers, limiting their performance in OD. To mitigate these issues, the Mean Squared Error (MSE) and Negative Logarithmic Likelihood (NLL) were firstly analyzed, and the importance of incorporating aleatoric uncertainty to AE-based OD was elucidated. Then the Weighted Negative Logarithmic Likelihood (WNLL) was proposed to adjust for the effect of uncertainty for different OD scenarios. Moreover, the Mean-Shift Scoring (MSS) method was proposed to utilize the local relationship of data to reduce the issue of false inliers caused by AE. Experiments on 32 real-world OD datasets proved the effectiveness of the proposed methods. The combination of WNLL and MSS achieved 41% relative performance improvement compared to the best baseline. In addition, MSS improved the detection performance of multiple AE-based outlier detectors by an average of 20%. The proposed methods have the potential to advance AE's development in OD.