Failure Mode

Failure mode analysis focuses on identifying and understanding how systems or processes can fail, aiming to improve reliability and safety. Current research emphasizes developing methods to predict and mitigate failures across diverse domains, including manufacturing, AI systems, and material science, employing techniques like reinforcement learning, Bayesian inference, and deep learning models (e.g., Support Vector Machines, physics-informed neural networks). This work is crucial for enhancing the robustness and trustworthiness of complex systems, impacting fields ranging from industrial automation and healthcare to the development of safe and reliable AI.

Papers

March 25, 2022

Common Failure Modes of Subcluster-based Sampling in Dirichlet Process Gaussian Mixture Models -- and a Deep-learning Solution
Vlad Winter, Or Dinari, Oren Freifeld
Deep Learning Data Clustering Failure Mode Random Initialization Based Sampling Dirichlet Process Gaussian Mixture

November 26, 2021

Failure Modes of Domain Generalization Algorithms
Tigran Galstyan, Hrayr Harutyunyan, Hrant Khachatrian, Greg Ver Steeg, Aram Galstyan
Domain Generalization Domain Invariant Representation Training Domain Failure Mode Domain Invariance

October 29, 2020

Understanding the Failure Modes of Out-of-Distribution Generalization
Vaishnavh Nagarajan, Anders Andreassen, Behnam Neyshabur
Machine Learning Model Distribution Generalization Spurious Correlation Linear Classifier Image Classification Datasets Failure Mode

Failure Mode

Papers

Common Failure Modes of Subcluster-based Sampling in Dirichlet Process Gaussian Mixture Models -- and a Deep-learning Solution

Failure Modes of Domain Generalization Algorithms

Understanding the Failure Modes of Out-of-Distribution Generalization