International audienceMotivated by applications in statistical inference, we propose two versions of a Stochastic Proximal Gradient algorithm for the maximization of composite functions. We establish the convergence in the concave case, when the Monte Carlo approximation is biased and the bias does not vanish along iterations.Motivés par des applications en inférence statistique, nous proposons deux versions stochastiques de l'algorithme Gradient Proximal pour la maximisation de fonctions composites. Nous établissons la convergence dans le cas concave, lorsque les approximations Monte Carlo sont biaisées avec un biais qui ne s'atténue pas au cours des itérations
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
Abstract. We study a stochastic version of the proximal gradient algorithm where the gradient is ana...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
Abstract We study the extension of the proximal gradient algorithm where only a stochastic gradient...
We study the extension of the proximal gradient algorithm where only a stochastic gradient estimate ...
Abstract. We study a stochastic version of the proximal gradient algorithm where the gradient is ana...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
Abstract. We study a stochastic version of the proximal gradient algorithm where the gradient is ana...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
International audienceMotivated by applications in statistical inference, we propose two versions of...
Abstract We study the extension of the proximal gradient algorithm where only a stochastic gradient...
We study the extension of the proximal gradient algorithm where only a stochastic gradient estimate ...
Abstract. We study a stochastic version of the proximal gradient algorithm where the gradient is ana...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
International audienceMotivated by penalized likelihood maximization in complex models, we study opt...
Abstract. We study a stochastic version of the proximal gradient algorithm where the gradient is ana...