From Compressive Clustering to Compressive Learning
نویسنده
چکیده
It is often useful to fit a probability model to a data collection, in order to concisely represent the data, to feed learning algorithms that work on densities, to extract features or, simply, to uncover underlying structures. A particularly popular probability model is the Gaussian Mixture Model (GMM). Among many other applications, GMM form a central tool to build time-frequency models of audio data that are used for audio source separation [2], and is traditionally fitted through the Expectation-Maximization (EM) algorithm [3]. However, when the collection is voluminous, memory and computation time can be prohibitive.
منابع مشابه
Compressive Statistical Learning with Random Feature Moments
We describe a general framework –compressive statistical learning– for resource-efficient largescale learning: the training collection is compressed in one pass into a low-dimensional sketch (a vector of random empirical generalized moments) that captures the information relevant to the considered learning task. A near-minimizer of the risk is computed from the sketch through the solution of a ...
متن کاملEstimating the Concrete Compressive Strength Using Hard Clustering and Fuzzy Clustering Based Regression Techniques
Understanding of the compressive strength of concrete is important for activities like construction arrangement, prestressing operations, and proportioning new mixtures and for the quality assurance. Regression techniques are most widely used for prediction tasks where relationship between the independent variables and dependent (prediction) variable is identified. The accuracy of the regressio...
متن کاملFast Approximate Text Document Clustering Using Compressive Sampling
Document clustering involves repetitive scanning of a document set, therefore as the size of the set increases, the time required for the clustering task increases and may even become impossible due to computational constraints. Compressive sampling is a feature sampling technique that allows us to perfectly reconstruct a vector from a small number of samples, provided that the vector is sparse...
متن کاملCompressive Spectral Clustering - Error Analysis
Compressive spectral clustering combines the distance preserving measurements of compressed sensing with the power of spectral clustering. Our analysis provides rigorous bounds on how small errors in the affinity matrix can affect the spectral coordinates and clusterability. This work generalizes the current perturbation results of two-class spectral clustering to incorporate multiclass cluster...
متن کاملAn Investigation of the Effect of Freezing on Strength and Durability of Dimension Stones Using Fuzzy Clustering Technique and Statistical Analysis
Western and North-Western regions of Iran feature very cold winters, a lot of snow, and freezing temperatures during most nights in December, January, February, and March. This directly influences the selection and applications of dimension stones in these areas. Freezing influences both mechanical and physical properties of rocks. Therefore, measuring the changes in values of these parameters ...
متن کاملEVELOPMENT OF ANFIS-PSO, SVR-PSO, AND ANN-PSO HYBRID INTELLIGENT MODELS FOR PREDICTING THE COMPRESSIVE STRENGTH OF CONCRETE
Concrete is the second most consumed material after water and the most widely used construction material in the world. The compressive strength of concrete is one of its most important mechanical properties, which highly depends on its mix design. The present study uses the intelligent methods with instance-based learning ability to predict the compressive strength of concrete. To achieve this ...
متن کامل