Failing Conceptually: Concept-Based Explanations of Dataset Shift

by   Maleakhi A. Wijaya, et al.
University of Cambridge

Despite their remarkable performance on a wide range of visual tasks, machine learning technologies often succumb to data distribution shifts. Consequently, a range of recent work explores techniques for detecting these shifts. Unfortunately, current techniques offer no explanations about what triggers the detection of shifts, thus limiting their utility to provide actionable insights. In this work, we present Concept Bottleneck Shift Detection (CBSD): a novel explainable shift detection method. CBSD provides explanations by identifying and ranking the degree to which high-level human-understandable concepts are affected by shifts. Using two case studies (dSprites and 3dshapes), we demonstrate how CBSD can accurately detect underlying concepts that are affected by shifts and achieve higher detection accuracy compared to state-of-the-art shift detection methods.


Interpretable Distribution Shift Detection using Optimal Transport

We propose a method to identify and characterize distribution shifts in ...

Data+Shift: Supporting visual investigation of data distribution shifts by data scientists

Machine learning on data streams is increasingly more present in multipl...

Preventing dataset shift from breaking machine-learning biomarkers

Machine learning brings the hope of finding new biomarkers extracted fro...

Deep Hypothesis Tests Detect Clinically Relevant Subgroup Shifts in Medical Images

Distribution shifts remain a fundamental problem for the safe applicatio...

Rectifying Group Irregularities in Explanations for Distribution Shift

It is well-known that real-world changes constituting distribution shift...

Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift

We might hope that when faced with unexpected inputs, well-designed soft...

Towards Practicable Sequential Shift Detectors

There is a growing awareness of the harmful effects of distribution shif...

Please sign up or login with your details

Forgot password? Click here to reset