156 citations · 361 across the 32 of their papers we have counts for
38 papers
Any Deep ReLU Network is Shallow
Mattia Jacopo Villani, Nandi Schoots
We constructively prove that every deep ReLU network can be rewritten as a functionally identical three-layer network with weights valued in the extended reals. Based on this proof…
Unwrapping All ReLU Networks
Mattia Jacopo Villani, Peter McBurney
Deep ReLU Networks can be decomposed into a collection of linear models, each defined in a region of a partition of the input space. This paper provides three results extending thi…
Feature Importance for Time Series Data: Improving KernelSHAP
Mattia Villani, Joshua Lockhart, Daniele Magazzeni
Feature importance techniques have enjoyed widespread attention in the explainable AI literature as a means of determining how trained machine learning models make their prediction…
Bayesian Prediction with Covariates Subject to Detection Limits
Caroline Svahn, Mattias Villani
Missing values in covariates due to censoring by signal interference or lack of sensitivity in the measuring devices are common in industrial problems. We propose a full Bayesian s…
Bayesian Modeling of Effective and Functional Brain Connectivity using Hierarchical Vector Autoregressions
Bertil Wegmann, Anders Lundquist, Anders Eklund +1
Analysis of brain connectivity is important for understanding how information is processed by the brain. We propose a novel Bayesian vector autoregression (VAR) hierarchical model…
Local Prediction Pools
Oscar Oelrich, Mattias Villani, Sebastian Ankargren
We propose local prediction pools as a method for combining the predictive distributions of a set of experts conditional on a set of variables believed to be related to the predict…
Dynamic Mixture of Experts Models for Online Prediction
Parfait Munezero, Mattias Villani, Robert Kohn
A mixture of experts models the conditional density of a response variable using a mixture of regression models with covariate-dependent mixture weights. We extend the finite mixtu…
Including topology change in Loop Quantum Gravity with topspin network formalism with application to homogeneous and isotropic cosmology
Mattia Villani
We apply topspin network formalism to Loop Quantum Gravity in order to include in the theory the possibility of changes in the topology of spacetime. We apply this formalism to thr…
Robust Real-Time Delay Predictions in a Network of High-Frequency Urban Buses
Hector Rodriguez-Deniz, Mattias Villani
Providing transport users and operators with accurate forecasts on travel times is challenging due to a highly stochastic traffic environment. Public transport users are particular…
Spectral Subsampling MCMC for Stationary Multivariate Time Series with Applications to Vector ARTFIMA Processes
Mattias Villani, Matias Quiroz, Robert Kohn +1
Spectral subsampling MCMC was recently proposed to speed up Markov chain Monte Carlo (MCMC) for long stationary univariate time series by subsampling periodogram observations in th…
Low-energy electromagnetic processes affecting free-falling test-mass charging for LISA and future space interferometers
Catia Grimani, Andrea Cesarini, Michele Fabi +1
Galactic cosmic rays and solar energetic particles charge gold-platinum, free-falling test masses (TMs) on board interferometers for the detection of gravitational waves in space.…
Recurrent galactic cosmic-ray flux modulation in L1 and geomagnetic activity during the declining phase of the solar cycle 24
Catia Grimani, Andrea Cesarini, Michele Fabi +3
Galactic cosmic-ray (GCR) flux short-term variations (1 month) in the inner heliosphere are mainly associated with the passage of high-speed solar wind streams (HSS) and interpl…
Bayesian Optimization of Hyperparameters from Noisy Marginal Likelihood Estimates
Oskar Gustafsson, Mattias Villani, Pär Stockhammar
Bayesian models often involve a small set of hyperparameters determined by maximizing the marginal likelihood. Bayesian optimization is a popular iterative method where a Gaussian…
When are Bayesian model probabilities overconfident?
Oscar Oelrich, Shutong Ding, Måns Magnusson +2
Bayesian model comparison is often based on the posterior distribution over the set of compared models. This distribution is often observed to concentrate on a single model even wh…
A Multilayered Block Network Model to Forecast Large Dynamic Transportation Graphs: an Application to US Air Transport
Hector Rodriguez-Deniz, Mattias Villani, Augusto Voltes-Dorta
Dynamic transportation networks have been analyzed for years by means of static graph-based indicators in order to study the temporal evolution of relevant network components, and…
Anatomically informed Bayesian spatial priors for fMRI analysis
David Abramian, Per Sidén, Hans Knutsson +2
Existing Bayesian spatial priors for functional magnetic resonance imaging (fMRI) data correspond to stationary isotropic smoothing filters that may oversmooth at anatomical bounda…
Spectral Subsampling MCMC for Stationary Time Series
Robert Salomone, Matias Quiroz, Robert Kohn +2
Bayesian inference using Markov Chain Monte Carlo (MCMC) on large datasets has developed rapidly in recent years. However, the underlying methods are generally limited to relativel…
Spatial 3D Matérn priors for fast whole-brain fMRI analysis
Per Sidén, Finn Lindgren, David Bolin +2
Bayesian whole-brain functional magnetic resonance imaging (fMRI) analysis with three-dimensional spatial smoothing priors has been shown to produce state-of-the-art activity maps…
Real-Time Robotic Search using Hierarchical Spatial Point Processes
Olov Andersson, Per Sidén, Johan Dahlin +2
Aerial robots hold great potential for aiding Search and Rescue (SAR) efforts over large areas. Traditional approaches typically searches an area exhaustively, thereby ignoring tha…
Modeling Text Complexity using a Multi-Scale Probit
Johan Falkenjack, Mattias Villani, Arne Jönsson
We present a novel model for text complexity analysis which can be fitted to ordered categorical data measured on multiple scales, e.g. a corpus with binary responses mixed with a…
Subsampling MCMC - An introduction for the survey statistician
Matias Quiroz, Mattias Villani, Robert Kohn +2
The rapid development of computing power and efficient Markov Chain Monte Carlo (MCMC) simulation algorithms have revolutionized Bayesian statistics, making it a highly practical i…
Physiological Gaussian Process Priors for the Hemodynamics in fMRI Analysis
Josef Wilzén, Anders Eklund, Mattias Villani
Background: Inference from fMRI data faces the challenge that the hemodynamic system that relates neural activity to the observed BOLD fMRI signal is unknown. New Method: We propos…
Hamiltonian Monte Carlo with Energy Conserving Subsampling
Khue-Dung Dang, Matias Quiroz, Robert Kohn +2
Hamiltonian Monte Carlo (HMC) samples efficiently from high-dimensional posterior distributions with proposed parameter draws obtained by iterating on a discretized version of the…
Efficient Covariance Approximations for Large Sparse Precision Matrices
Per Sidén, Finn Lindgren, David Bolin +1
The use of sparse precision (inverse covariance) matrices has become popular because they allow for efficient algorithms for joint inference in high-dimensional models. Many applic…