Concept
Conjugate prior
Related publications (127)
ZigZag: Universal Sampling-free Uncertainty Estimation Through Two-Step Inference
Nikita Durasov, Minh Hieu Lê, Nik Joel Dorndorf
Whereas the ability of deep networks to produce useful predictions on many kinds of data has been amply demonstrated, estimating the reliability of these predictions remains challenging. Sampling approaches such as MC-Dropout and Deep Ensembles have emerge ...
2024It’s All Relative: Learning Interpretable Models for Scoring Subjective Bias in Documents from Pairwise Comparisons
Matthias Grossglauser, Aswin Suresh, Chi Hsuan Wu
We propose an interpretable model to score the subjective bias present in documents, based only on their textual content. Our model is trained on pairs of revisions of the same Wikipedia article, where one version is more biased than the other. Although pr ...
2024Toward Reliable Human Pose Forecasting With Uncertainty
Alexandre Massoud Alahi, Saeed Saadatnejad, Taylor Ferdinand Mordan, Parham Saremi
Recently, there has been an arms race of pose forecasting methods aimed at solving the spatio-temporal task of predicting a sequence of future 3D poses of a person given a sequence of past observed ones. However, the lack of unified benchmarks and limited ...
2024Methodology for selecting measurement points that optimize information gain for model updating
Information collected through sensor measurements has the potential to improve knowledge of complex-system behavior, leading to better decisions related to system management. In this situation, and particularly when using digital twins, the quality of sens ...
2023Economic Effect of Patents and the Patent System: Insights from Linked Patent-Product Data
This thesis investigates the economic effect of patents and the patent system through the lens of patent commercialisation. The thesis is composed of four chapters, where each chapter is an independent scientific paper. In the first chapter, we present a n ...
EPFL2023Fast Bayesian estimation of spatial count data models
Spatial count data models are used to explain and predict the frequency of phenomena such as traffic accidents in geographically distinct entities such as census tracts or road segments. These models are typically estimated using Bayesian Markov chain Mont ...
2021Learning, compression, and leakage: Minimising classification error via meta-universal compression principles
Michael Christoph Gastpar, Fernando Rosas de Andraca
Learning and compression are driven by the common aim of identifying and exploiting statistical regularities in data, which opens the door for fertile collaboration between these areas. A promising group of compression techniques for learning scenarios is ...
IEEE2021Learning in Volatile Environments With the Bayes Factor Surprise
Wulfram Gerstner, Johanni Michael Brea, Alireza Modirshanechi, Vasiliki Liakoni
Surprise-based learning allows agents to rapidly adapt to nonstationary stochastic environments characterized by sudden changes. We show that exact Bayesian inference in a hierarchical model gives rise to a surprise-modulated trade-off between forgetting o ...
MIT PRESS2021DiffuserCam Project
William Cappelletti, Jonathan Philippe Reymond, Matteo Pariset
We test the reconstruction power of a lensless cam- era on a custom dataset. We define a linear inverse problem on the raw image, based on a PSF estimation of the the sensing device and the actual captures. We implement various regularizations to enforce s ...
2021The Impact of Changes in Resolution on the Persistent Homology of Images
Digital images enable quantitative analysis of material properties at micro and macro length scales, but choosing an appropriate resolution when acquiring the image is challenging. A high resolution means longer image acquisition and larger data requiremen ...
IEEE2021