用于紧凑潜在表示中无监督异常值检测的空缺孔洞方法
Vacant Holes for Unsupervised Detection of the Outliers in Compact Latent Representation
AI总结：
针对经典变分自编码器（VAE）的理论缺陷与异常检测难题，通过亚历山德罗夫延拓和固定利普希茨常数实现潜在空间紧凑化，发现异常输入易落入空间空缺孔洞，设计专用检测分数，在多基准上取得良好无监督异常检测效果。
摘要：
异常值检测对于任何在现实世界中部署运行的机器学习模型而言都至关重要。对于已被证明会对这类输入产生过度置信的深度神经网络来说，这一任务尤为关键。此外，即便能够估计输入概率密度的深度生成模型，也难以完成这项任务。在本研究中，我们聚焦于这类模型中的特定类型：变分自编码器（Variational Autoencoders, VAEs）。首先，我们揭示了经典VAE模型假设中存在的一个重大理论缺陷。其次，我们为深度神经网络映射到潜在空间的像施加了一种适配的拓扑性质：紧凑性，以缓解该缺陷，并通过将内点和异常点挤压在一起，获得可证明地将像限定在确定范围内的手段。我们通过两种方法实现紧凑性：（i）亚历山德罗夫延拓（Alexandroff extension），以及（ii）在VAE编码器的映射上设置固定的利普希茨连续性常数。最后且最重要的是，我们发现异常输入主要倾向于落在紧凑空间内的空缺潜在孔洞中，这使得它们能够被成功识别。基于此，我们提出了一种专门设计的孔洞检测分数，并在多个基线基准上对该方案进行了评估，取得了颇具前景的结果。
英文摘要：
Detection of the outliers is pivotal for any machine learning model deployed and operated in real-world. It is essential for the Deep Neural Networks that were shown to be overconfident with such inputs. Moreover, even deep generative models that allow estimation of the probability density of the input fail in achieving this task. In this work, we concentrate on the specific type of these models: Variational Autoencoders (VAEs). First, we unveil a significant theoretical flaw in the assumption of the classical VAE model. Second, we enforce an accommodating topological property to the image of the deep neural mapping to the latent space: compactness to alleviate the flaw and obtain the means to provably bound the image within the determined limits by squeezing both inliers and outliers together. We enforce compactness using two approaches: (i) Alexandroff extension and (ii) fixed Lipschitz continuity constant on the mapping of the encoder of the VAEs. Finally and most importantly, we discover that the anomalous inputs predominantly tend to land on the vacant latent holes within the compact space, enabling their successful identification. For that reason, we introduce a specifically devised score for hole detection and evaluate the solution against several baseline benchmarks achieving promising results.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
