[Seminar]MLDS Seminar 2023-4 by Mr. Tobias Freidling (University of Cambridge), Dr. Mohammad Sabokrou (Staff Scientist, OIST), Seminar Room L5D23

Speaker 1: Mr. Tobias Freidling, Ph. D. Student, University of Cambridge

Title: Sensitivity Analysis with the R^2-calculus

Abstract: Causal inference necessarily relies upon untestable identification assumptions; hence, it is crucial to assess the robustness of obtained results to potential violations. However, such sensitivity analysis is only occasionally undertaken in practice as many existing methods only apply to relatively simple models and their results are often difficult to interpret. We take a more flexible approach to sensitivity analysis and view it as a constrained stochastic optimization problem. This work focuses on linear models with an unmeasured confounder and a potential instrument. In this setting, the R^2-calculus – a set of algebraic rules that relates different (partial) R^2-values and correlations – emerges as the key tool for sensitivity analysis. It can be applied to identify the bias of the family of k-class estimators, which includes the OLS and TSLS estimators, as well as construct sensitivity models flexibly. For instance, practitioners can specify their assumptions on the unmeasured confounder by comparing
its influence on treatment/outcome with an observed variable. We further address the problem of constructing sensitivity intervals using a bootstrap approach. We illustrate the proposed methods with a real data example and provide user-friendly visualization tools.

Speaker 2: Dr. Mohammad Sabokrou, Staff Scientist, OIST 

Title: Deep Learning Advancements in Anomaly Detection for Computer Vision

Abstract: This talk focuses on the advancements in anomaly detection in computer vision and image processing over the past five years, specifically using deep learning approaches. It discusses the main methods developed for anomaly detection and clarifies the similarities and differences between related fields, such as novelty detection, open set recognition, and OOD detection. The talk also addresses misconceptions within the literature and highlights the need to correct certain anomaly detection aspects. It concludes by discussing the current challenges and open problems in the field. Overall, attendees will gain insights into recent developments, understand the distinctions between related fields, and be aware of the existing challenges and opportunities in anomaly detection.