Summary of Retrieval Augmented Deep Anomaly Detection For Tabular Data, by Hugo Thimonier et al.
Retrieval Augmented Deep Anomaly Detection for Tabular Data
by Hugo Thimonier, Fabrice Popineau, Arpad Rimmel, Bich-Liên Doan
First submitted to arxiv on: 30 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper investigates the use of retrieval-augmented models for anomaly detection in tabular data, a challenging task despite deep learning’s success with unstructured data. The authors propose a reconstruction-based approach where a transformer model learns to reconstruct masked features of normal samples, leveraging KNN-based and attention-based modules to select relevant samples for help. Experimental results on 31 benchmark datasets show that incorporating retrieval modules significantly boosts performance, supporting the idea that retrieval-augmented models can enhance anomaly detection on tabular data. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about using special kinds of artificial intelligence (AI) to find unusual patterns in tables of numbers and words. Right now, these AI models are great at finding patterns in unorganized data like pictures or sounds. But when it comes to organized data like spreadsheets, they don’t do as well. The researchers tried a new way to use these AI models by having them look at examples of normal data and then trying to recreate the missing parts. They also experimented with different ways for the model to choose which examples to use. In their tests on 31 different datasets, they found that this new approach worked really well and can help us find unusual patterns in spreadsheets. |
Keywords
* Artificial intelligence * Anomaly detection * Attention * Deep learning * Transformer
