English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Autoencoders were widely used in many machine learning tasks thanks to theirstrong learning ability which has drawn great interest among researchers in thefield of outlier detection. However, conventional autoencoder-based methodslacked considerations in two aspects. This limited their performance in outlierdetection. First, the mean squared error used in conventional autoencodersignored the judgment uncertainty of the autoencoder, which limited theirrepresentation ability. Second, autoencoders suffered from the abnormalreconstruction problem: some outliers can be unexpectedly reconstructed well,making them difficult to identify from the inliers. To mitigate theaforementioned issues, two novel methods were proposed in this paper. First, anovel loss function named Probabilistic Reconstruction Error (PRE) wasconstructed to factor in both reconstruction bias and judgment uncertainty. Tofurther control the trade-off of these two factors, two weights were introducedin PRE producing Adjustable Probabilistic Reconstruction Error (APRE), whichbenefited the outlier detection in different applications. Second, aconceptually new outlier scoring method based on mean-shift (MSS) was proposedto reduce the false inliers caused by the autoencoder. Experiments on 32real-world outlier detection datasets proved the effectiveness of the proposedmethods. The combination of the proposed methods achieved 41% of the relativeperformance improvement compared to the best baseline. The MSS improved theperformance of multiple autoencoder-based outlier detectors by an average of20%. The proposed two methods have the potential to advance autoencoder'sdevelopment in outlier detection. The code is available on www.OutlierNet.comfor reproducibility.

Improving Autoencoder-based Outlier Detection with Adjustable Probabilistic Reconstruction Error and Mean-shift Outlier Scoring