Probability Sliding Curves and Probabilistic Graphs – We present a Bayesian approach to sparse convex optimization by exploiting the similarity of the coefficients of the two discrete sets. Our approach combines a Bayesian formulation with a logistic regression technique and an approximate posterior estimator by means of a conditional Bayesian inference algorithm. We show that this can be efficiently computed from the posterior estimator and are able to perform well, thanks to the use of a Bayesian procedure. Our results imply that the Bayesian technique is a valid method for sparsity constrained convex optimization, in which the approximation of the posterior estimator is a condition which can be fulfilled by the posterior estimator.

Convolutional Neural Networks aims at using a large amount of labelled information (the labeled data) to efficiently interpret semantic patterns, such as images with varying orientations. We propose to use deep recurrent neural networks (RNNs) for this task by using contextual tasks to learn and process labels of images. Firstly, a convolutional neural network is connected to the convolutional layers of the RNN for this task. Then, an RNN can learn to infer the contextual semantic patterns, and then use them to perform image-level task based on the contextual labels. We validate our approach on a dataset of images that exhibit a variety of orientations and labels, and show that it is able to interpret the labels better than other models trained to discriminate between orientations and labels.

Training of Convolutional Neural Networks

Training a Sparse Convolutional Neural Network for Receptive Field Detection

# Probability Sliding Curves and Probabilistic Graphs

Modeling and Analysis of Non-Uniform Graphical Models as Bayesian Models

A Bayesian Network Architecture for Multi-Modal Image Search, Using Contextual TasksConvolutional Neural Networks aims at using a large amount of labelled information (the labeled data) to efficiently interpret semantic patterns, such as images with varying orientations. We propose to use deep recurrent neural networks (RNNs) for this task by using contextual tasks to learn and process labels of images. Firstly, a convolutional neural network is connected to the convolutional layers of the RNN for this task. Then, an RNN can learn to infer the contextual semantic patterns, and then use them to perform image-level task based on the contextual labels. We validate our approach on a dataset of images that exhibit a variety of orientations and labels, and show that it is able to interpret the labels better than other models trained to discriminate between orientations and labels.