Neural Fisher Discriminant Analysis – Neural network models contain two main components, classification and segmentation, which are very similar but which are not easily distinguishable. Classifying the network structure can be tedious and time consuming, especially for large networks. This work tackles the task of classifying a large set of MNIST digits using neural networks (NN). We first propose a neural network model of MNIST digits which has a multi-layer perceptron for classification. Then we apply a neural network to classify MNIST digits using a multi-task learning algorithm. Experimental results demonstrate that the proposed model outperforms the state-of-the-art MNIST digits classification method.
We propose a new method of learning a representation from music. It is based on a notion of the melody and the rhythm, which provides a direct interpretation of the music. We formulate the algorithm as a neural-network learning and we prove the relevance for a music classification task. We show that we can learn the melody of a song by learning the rhythm of the song as the melody of the song. We describe the algorithm and the experiments it demonstrates on a challenging music classification task.
Fast Partition Learning for Partially Observed Graphs
A General Algorithm for Grouping Visual Features into Semantic Spaces
Neural Fisher Discriminant Analysis
A Novel Hybrid Model for Computing Pairwise Pairwise Markov Forests
A comparative study of different types of recurrent neural networks for music classificationWe propose a new method of learning a representation from music. It is based on a notion of the melody and the rhythm, which provides a direct interpretation of the music. We formulate the algorithm as a neural-network learning and we prove the relevance for a music classification task. We show that we can learn the melody of a song by learning the rhythm of the song as the melody of the song. We describe the algorithm and the experiments it demonstrates on a challenging music classification task.