Summary of Generating Realistic Adversarial Examples For Business Processes Using Variational Autoencoders, by Alexander Stevens et al.
Generating Realistic Adversarial Examples for Business Processes using Variational Autoencoders
by Alexander Stevens, Jari Peeperkorn, Johannes De Smedt, Jochen De Weerdt
First submitted to arxiv on: 21 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 addresses a significant vulnerability in predictive process monitoring, where minor changes to activity sequences can lead to incorrect predictions due to underlying constraints. To address this challenge, the authors propose two novel latent space attacks that generate realistic adversaries tailored to the business process context. Unlike imperceptible pixel-level changes in computer vision, these attacks add noise to the latent space representation of input data. The methods are domain-agnostic and do not rely on process-specific knowledge, instead restricting adversarial examples to learned class-specific data distributions. Evaluation is conducted on 11 real-life event logs and four predictive models using seven different attacking methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps keep business processes safe by making sure computer systems don’t get tricked into thinking something is happening when it’s not. It does this by creating special kinds of fake data that are hard to spot, but still realistic. The authors come up with two new ways to create these fake data points, which can be used to test how well business process monitoring systems work. They tested their methods on real-life data from 11 different companies and four different prediction models. |
Keywords
» Artificial intelligence » Latent space
