We consider a regression problem where uncertainty affects to the dependent variable of the elements of the database. A model based on the standard -Support Vector Regression approach is given, where two hyperplanes need to be constructed to predict the interval-valued dependent variable. By using the Hausdorff distance to measure the error between predicted and real intervals, a convex quadrat...

Abstract: Training a SVR (support vector regression) requires the solution of a very large QP (quadratic programming) optimization problem. Despite the fact that this type of problem is well understood, the existing training algorithms are very complex and slow. In order to solve these problems, this paper firstly introduces a new way to make SVR have the similar mathematic form as that of a su...

Forecasting the number of warranty claims is vitally important for manufacturers/warranty providers in preparing fiscal plans. In existing literature, a number of techniques such as log-linear Poisson models, Kalman filter, time series models, and artificial neural network models have been developed. Nevertheless, one might find two weaknesses existing in these approaches: (1) they do not consi...

In this paper we describe a novel extension of the support vector machine, called the deep support vector machine (DSVM). The original SVM has a single layer with kernel functions and is therefore a shallow model. The DSVM can use an arbitrary number of layers, in which lower-level layers contain support vector machines that learn to extract relevant features from the input patterns or from the...

We present a novel algorithm for sparse online greedy kernelbased nonlinear regression. This algorithm improves current approaches to kernel-based regression in two aspects. First, it operates online at each time step it observes a single new input sample, performs an update and discards it. Second, the solution maintained is extremely sparse. This is achieved by an explicit greedy sparsificati...

The ever-growing volume and value of data has raised increasing pressure for long-term data protection in storage systems. Moreover, the redundancy in data further aggravates such pressure in these systems. It has become a serious problem to protect data while eliminating data redundancy, saving storage space and network bandwidth as well. Data deduplication techniques greatly optimize storage ...

In a remote backup system, transaction processing takes place at the primary and the log records generated at the primary are propagated to the remote backup which uses them to reconstruct a recent state of the database at the primary. In the event of a primary failure, the backup system takes over transaction processing without causing users to observe a breach in service. Existing remote back...

We consider the problem of deriving class-size independent generalization bounds for some regularized discriminative multi-category classification methods. In particular, we obtain an expected generalization bound for a standard formulation of multi-category support vector machines. Based on the theoretical result, we argue that the formulation over-penalizes misclassification error, which in t...

This paper aims to decrease the time complexity of multi-output relevance vector regression from O ( VM ) to O ( V 3 +M ) , where V is the number of output dimensions, M is the number of basis functions, and V < M . The experimental results demonstrate that the proposed method is more competitive than the existing method, with regard to computation time. MATLAB codes are available at http://www...

This paper describes an on-line method for building ε-insensitive support vector machines for regression as described in (Vapnik, 1995). The method is an extension of the method developed by (Cauwenberghs & Poggio, 2000) for building incremental support vector machines for classification. Machines obtained by using this approach are equivalent to the ones obtained by applying exact methods like...