Statistical learning theory

About: Statistical learning theory is a research topic. Over the lifetime, 1618 publications have been published within this topic receiving 158033 citations.

Application of SVM in Predicting the Strength of Cement Stabilized Soil

Abstract: Support Vector Machine is a powerful machine learning technique based on statistical learning theory. This paper investigates the potential of support vector machines based regression approach to model the strength of cement stabilized soil from test dates. Support Vector Machine model is proposed to predict compressive strength of cement stabilized soil. And the effects of selecting kernel function on Support Vector Machine modeling are also analyzed. The results show that the Support Vector Machine is more precise in measuring the strength of cement than traditional methods. The Support Vector Machine method has advantages in its simple structure，excellent capability in studying and good application prospects, also it provide us with a novel method of measuring the strength of cement stabilized soil.

Contributions to statistical learning in sparse models

Abstract: The aim of this habilitation thesis is to give an overview of my works on high-dimensional statistics and statistical learning, under various sparsity assumptions. In a first part, I will describe the major challenges of high-dimensional statistics in the context of the generic linear regression model. After a brief review of existing results, I will present the theoretical study of aggregated estimators that was done in (Alquier & Lounici 2011). The second part essentially aims at providing extensions of the various theories presented in the first part to the estimation of time series models (Alquier & Doukhan 2011, Alquier & Wintenberger 2013, Alquier & Li 2012, Alquier, Wintenberger & Li 2012). Finally, the third part presents various extensions to nonparametric models, or to specific applications such as quantum statistics (Alquier & Biau 2013, Guedj & Alquier 2013, Alquier, Meziani & Peyre 2013, Alquier, Butucea, Hebiri, Meziani & Morimae 2013, Alquier 2013, Alquier 2008). In each section, we provide explicitely the estimators used and, as much as possible, optimal oracle inequalities satisfied by these estimators.

A study of applying support vector machine and Genetic Algorithm to software reliability forecasting

Abstract: Software reliability prediction models is very helpful for developers and testers to know the phase in which corrective action need to be performed in order to achieve target reliability estimate. In this paper, an SVM-based model for software reliability forecasting is proposed. Support vector machine (SVM) is a new method based on statistical learning theory. It has been successfully used to solve nonlinear regression and time series problems. However, SVM has rarely been applied to software reliability prediction. In addition, the parameters of SVM are determined by Genetic Algorithm (GA). It is also demonstrated that only recent failure data is enough for model training. This feature that the model does not use all available failure data enables software developers and testers to obtain general ideas about software reliability in the early phase of testing process. Two types of model input data selection in the literature are employed to illustrate the performances of various prediction models. Empirical results show that the proposed model is more precise in its reliability prediction and is less dependent on the size of failure data comparing with the other forecasting models.

Performance improvement in traffic vision systems using svms

Abstract: Computer vision system is one of important research topics in ITS(Intelligent Transport Systems). Moreover, Neural Networks have been increasingly and successfully applied to many problems for ITS. Even though there are currently many different types of neural network models, Backpropagation is the most popular neural network model. It is however known that the Support Vector Machines (SVMs) based on the statistical learning theory is currently another efficient approach for pattern recognition problem since their remarkable performance in terms of prediction accuracy. In this research, two different models, Backpropagation and SVMs have been studied to compare their performance in predictive accuracy through the experiment with real world image data of traffic scenes. Experimental results show that SVMs can provide higher performance in terms of prediction performance than any other models.