Technical reports 

About: Technical reports is an academic journal. The journal publishes majorly in the area(s): Estimator & Support vector machine. Over the lifetime, 576 publications have been published receiving 12477 citations.

Making large scale SVM learning practical

Abstract: Training a support vector machine SVM leads to a quadratic optimization problem with bound constraints and one linear equality constraint Despite the fact that this type of problem is well understood, there are many issues to be considered in designing an SVM learner In particular, for large learning tasks with many training examples on the shelf optimization techniques for general quadratic programs quickly become intractable in their memory and time requirements SVM light is an implementation of an SVM learner which addresses the problem of large tasks This chapter presents algorithmic and computational results developed for SVM light V 20, which make large-scale SVM training more practical The results give guidelines for the application of SVMs to large domains

Strucchange: An R package for testing for structural change in linear regression models

Abstract: This paper introduces ideas and methods for testing for structural change in linear regression models and presents how these have been realized in an R package called strucchange. It features tests from the generalized uctuation test framework as well as from the F test (Chow test) framework. Extending standard signi_cance tests it contains methods to _t, plot and test empirical uctuation processes (like CUSUM, MOSUM and estimatesbased processes) on the one hand and to compute, plot and test sequences of F statistics with the supF, aveF and expF test on the other. Thus, it makes powerful tools available to display information about structural changes in regression relationships and to assess their signi_cance. Furthermore it is described how incoming data can be monitored online.

Testing and dating of structural changes in practice

Abstract: The paper presents an approach to the analysis of data that contains (multiple) structural changes in a linear regression setup. We implement various strategies which have been suggested in the literature for testing against structural changes as well as a dynamic programming algorithm for the dating of the breakpoints in the R statistical software package. Using historical data on Nile river discharges, road casualties in Great Britain and oil prices in Germany it is shown that changes in the mean of a time series as well as in the coefficients of a linear regression are easily matched with identifiable historical, political or economic events.

Logics of Communication and Change

Abstract: Current dynamic epistemic logics for analyzing effects of informational events often become cumbersome and opaque when common knowledge is added for groups of agents. Still, postconditions involving common knowledge are essential to successful multi-agent communication. We propose new systems that extend the epistemic base language with a new notion of ‘relativized common knowledge’, in such a way that the resulting full dynamic logic of information flow allows for a compositional analysis of all epistemic postconditions via perspicuous ‘reduction axioms’. We also show how such systems can deal with factual alteration, rather than just information change, making them cover a much wider range of realistic events. After a warm-up stage of analyzing logics for public announcements, our main technical results are expressivity and completeness theorems for a much richer logic that we call LCC. This is a dynamic epistemic logic whose static base is propositional dynamic logic (PDL), interpreted epistemically. This system is capable of expressing all model-shifting operations with finite action models, while providing a compositional analysis for a wide range of informational events. This makes LCC a serious candidate for a standard in dynamic epistemic logic, as we illustrate by analyzing some complex communication scenarios, including sending successive emails with both ‘cc’ and ‘bcc’ lines, and other private announcements to subgroups. Our proofs involve standard modal techniques, combined with a new application of Kleene’s theorem on finite automata, as well as new Ehrenfeucht games of model comparison. © 2006 Elsevier Inc. All rights reserved. ∗ Corresponding author. E-mail addresses: johan@science.uva.nl (J. van Benthem), jan.van.eijck@cwi.nl, jve@cwi.nl (J. van Eijck), B.P.Kooi@rug.nl (B. Kooi). 0890-5401/$ see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ic.2006.04.006 J. van Benthem et al. / Information and Computation 204 (2006) 1620–1662 1621

A Central Limit Theorem for Realised Power and Bipower Variations of Continuous Semimartingales

Abstract: Consider a semimartingale of the form Y_{t}=Y_0+\int _0^{t}a_{s}ds+\int _0^{t}_{s-} dW_{s}, where a is a locally bounded predictable process and (the "volatility") is an adapted right--continuous process with left limits and W is a Brownian motion. We define the realised bipower variation process V(Y;r,s)_{t}^n=n^{((r+s)/2)-1} \sum_{i=1}^{[nt]}|Y_{(i/n)}-Y_{((i-1)/n)}|^{r}|Y_{((i+1)/n)}-Y_{(i/n)}|^{s}, where r and s are nonnegative reals with r+s>0. We prove that V(Y;r,s)_{t}n converges locally uniformly in time, in probability, to a limiting process V(Y;r,s)_{t} (the "bipower variation process"). If further is a possibly discontinuous semimartingale driven by a Brownian motion which may be correlated with W and by a Poisson random measure, we prove a central limit theorem, in the sense that \sqrt(n) (V(Y;r,s)^n-V(Y;r,s)) converges in law to a process which is the stochastic integral with respect to some other Brownian motion W', which is independent of the driving terms of Y and \sigma. We also provide a multivariate version of these results.