Evaluation of a Multilayer Weighted Gaussian Process Latent Variable Model for Pattern Recognition – We present the first general-purpose, scalable and robust method to infer the structure of a deep neural network using only a small number of observations. Our method first partitions the input of a neural network by three layers. Then it is analyzed by a feature fusion technique guided by a novel method for representing the network structure. Finally, we propose a novel unsupervised learning scheme for inferring the network structure based on local feature representations of network features. Our approach leverages the ability of large, unsupervised feature datasets to form a model, and presents a fast learning algorithm that outperforms state-of-the-art unsupervised methods on various datasets.
The problem of segmentation of unstructured data, also known as unstructured sparse coding, involves solving a sparse coding problem which encodes a sequence of unstructured variables into a sparse coding block. Unstructured coding (NC) offers a fast solution for sparse coding problems with a fixed representation. In this paper, the representation of the input data is represented by a matrix, which is considered as a sparse coding matrix. In order to solve the sparse coding problem, we propose an efficient formulation of NC, which is based on a non-convex optimization problem. An algorithm for solving the NC problem is presented. Experiments on a variety of datasets show a significant reduction in the size and computation time (up to 10^10 m imes 10^8) compared with the classical NC, which uses data from multiple viewpoints and which requires to maintain a constant size matrix dimension. By minimizing the dimension of the matrix, the proposed algorithm is also able to obtain high accuracy results without significant computation overhead.
Semi-supervised salient object detection via joint semantic segmentation
An Analysis of Deep Learning for Time-Varying Brain Time Series Feature Classification
Evaluation of a Multilayer Weighted Gaussian Process Latent Variable Model for Pattern Recognition
Tensor Logistic Regression via Denoising Random Forest
Convex Sparsification of Unstructured Aggregated DataThe problem of segmentation of unstructured data, also known as unstructured sparse coding, involves solving a sparse coding problem which encodes a sequence of unstructured variables into a sparse coding block. Unstructured coding (NC) offers a fast solution for sparse coding problems with a fixed representation. In this paper, the representation of the input data is represented by a matrix, which is considered as a sparse coding matrix. In order to solve the sparse coding problem, we propose an efficient formulation of NC, which is based on a non-convex optimization problem. An algorithm for solving the NC problem is presented. Experiments on a variety of datasets show a significant reduction in the size and computation time (up to 10^10 m imes 10^8) compared with the classical NC, which uses data from multiple viewpoints and which requires to maintain a constant size matrix dimension. By minimizing the dimension of the matrix, the proposed algorithm is also able to obtain high accuracy results without significant computation overhead.