We consider the problem of parameter es-timation using weakly supervised datasets, where a training sample consists of the input and a partially specified annotation, which we refer to as the output. The missing in-formation in the annotation is modeled using latent variables. Previous methods overbur-den a single distribution with two separate tasks: (i) modeling the uncertainty in the la-tent variables during training; and (ii) mak-ing accurate predictions for the output and the latent variables during testing. We pro-pose a novel framework that separates the demands of the two tasks using two distribu-tions: (i) a conditional distribution to model the uncertainty of the latent variables for a given input-output pair; and (ii) a delta dis...
Uncertainty quantification has received increasing attention in machine learning in the recent past....
Deep learning tools have gained tremendous attention in applied machine learning. However such tools...
We suggest a general approach to quantification of different forms of aleatoric uncertainty in regre...
International audienceWe consider the problem of parameter estimation using weakly supervised datase...
International audienceWe consider the problem of parameter estimation using weakly supervised datase...
We consider the problem of learning the parameters of a structured output prediction model, that is,...
We consider the problem of learning the parameters of a structured output prediction model, that is,...
We consider the problem of learning the parameters of a structured output prediction model, that is,...
Probabilistic graphical models such as Markov random fields provide a powerful framework and tools f...
I conduct a systematic study of probabilistic latent variable models (PLVMs) with applications to kn...
Machine learning models automatically learn from historical data to predict unseen events. Such eve...
Abstract. Uncertainty of the input data is a common issue in machine learning. In this paper we show...
Machine learning models automatically learn from historical data to predict unseen events. Such even...
This dissertation covers a collection of supervised learning methods targeted to data with complex d...
Abstract—Active learning methods aim to choose the most informative instances to effectively learn a...
Uncertainty quantification has received increasing attention in machine learning in the recent past....
Deep learning tools have gained tremendous attention in applied machine learning. However such tools...
We suggest a general approach to quantification of different forms of aleatoric uncertainty in regre...
International audienceWe consider the problem of parameter estimation using weakly supervised datase...
International audienceWe consider the problem of parameter estimation using weakly supervised datase...
We consider the problem of learning the parameters of a structured output prediction model, that is,...
We consider the problem of learning the parameters of a structured output prediction model, that is,...
We consider the problem of learning the parameters of a structured output prediction model, that is,...
Probabilistic graphical models such as Markov random fields provide a powerful framework and tools f...
I conduct a systematic study of probabilistic latent variable models (PLVMs) with applications to kn...
Machine learning models automatically learn from historical data to predict unseen events. Such eve...
Abstract. Uncertainty of the input data is a common issue in machine learning. In this paper we show...
Machine learning models automatically learn from historical data to predict unseen events. Such even...
This dissertation covers a collection of supervised learning methods targeted to data with complex d...
Abstract—Active learning methods aim to choose the most informative instances to effectively learn a...
Uncertainty quantification has received increasing attention in machine learning in the recent past....
Deep learning tools have gained tremendous attention in applied machine learning. However such tools...
We suggest a general approach to quantification of different forms of aleatoric uncertainty in regre...