Learn More
In many applications, data appear with a huge number of instances as well as features. Linear Support Vector Machines (SVM) is one of the most popular tools to deal with such large-scale sparse data. This paper presents a novel dual coordinate descent method for linear SVM with L1-and L2-loss functions. The proposed method is simple and reaches an(More)
Efficient training of direct multi-class formulations of linear Support Vector Machines is very useful in applications such as text classification with a huge number examples as well as features. This paper presents a fast dual method for this training. The main idea is to sequentially traverse through the training set and optimize the dual variables(More)
Gaussian processes are powerful regression models specified by parameterized mean and covariance functions. Standard approaches to choose these parameters (known by the name hyperparameters) are maximum likelihood and maximum a posteriori. In this article, we propose and investigate predictive approaches based on Geisser's predictive sample reuse (PSR)(More)
Newly developed insulin-sensitizing agents, which target the nuclear receptor peroxisome proliferator-activated receptor-gamma have recently been appreciated to exhibit potent anti-inflammatory actions. Since stroke is associated with an intense inflammatory response, we reasoned that these agents may ameliorate injury from stroke. We report that(More)
Gaussian Processes (GPs) are promising Bayesian methods for classification and regression problems. They have also been used for semi-supervised learning tasks. In this paper, we propose a new algorithm for solving semi-supervised binary classification problem using sparse GP regression (GPR) models. It is closely related to semi-supervised learning based(More)
A recent comparison of various algorithms for sequence labeling by Nguyen and Guo indicated that SVM-struct has much superior generalization performance than CRFs. In this short report we point out that the above difference mainly arises because that comparison employed different softwares that use different internal feature functions. When the two methods(More)
Distributed training of l 1 regularized classifiers has received great attention recently. Most existing methods approach this problem by taking steps obtained from approximating the objective by a quadratic approximation that is decoupled at the individual variable level. These methods are designed for multicore and MPI platforms where communication costs(More)
Scalable machine learning over big data stored on a cluster of commodity machines with significant communication costs has become important in recent years. In this paper we give a novel approach to the distributed training of linear classi-fiers (involving smooth losses and L2 regularization) that is designed to reduce communication costs. At each(More)
Stroke is a devastating disease with limited treatment options. Recently, we found that the peroxisome proliferator-activated receptor-gamma (PPARgamma) agonists troglitazone and pioglitazone reduce injury and inflammation in a rat model of transient cerebral ischemia. The mechanism of this protection is unclear, as these agents can act through PPAR-gamma(More)
In many real world prediction problems the output is a struc-tured object like a sequence or a tree or a graph. Such problems range from natural language processing to computational biology or computer vision and have been tackled using algorithms, referred to as structured output learning algorithms. We consider the problem of structured classification. In(More)