Showing papers by "Amaury Lendasse published in 2004"
TL;DR: This paper compares two recently published methods for nonlinear projection: Isomap and Curvilinear Distance Analysis (CDA), which are based on a nonlinear variant of MDS called Curvil inear Component Analysis (CCA).
190 citations
Proceedings Article•
01 Jan 2004
TL;DR: This paper presents the CATS Benchmark and the results of the competition organised during the IJCNN’04 conference in Budapest, which was the prediction of 100 missing values divided into five groups of twenty consecutive values.
Abstract: This paper presents the CATS Benchmark and the results of the competition organised during the IJCNN’04 conference in Budapest. Twenty-four papers and predictions have been submitted and seventeen have been selected. The goal of the competition was the prediction of 100 missing values divided into five groups of twenty consecutive values.
96 citations
TL;DR: The proof of the stability of the method for long-term forecasting is given, as well as illustrations of the utilization of the methods both in the scalar and vectorial cases.
19 citations
Proceedings Article•
01 Jan 2004
TL;DR: This paper presents a simple procedure to obtain a fast approximation of this generalization error of nonlinear regression models with a reduced computation time.
Abstract: The Bootstrap resampling method may be efficiently used to estimate the generalization error of nonlinear regression models, as artificial neural networks and especially Least-square Support Vector Machines. Nevertheless, the use of the Bootstrap implies a high computational load. In this paper we present a simple procedure to obtain a fast approximation of this generalization error with a reduced computation time. This proposal is based on empirical evidence and included in a simulation procedure.
13 citations
25 Jul 2004
TL;DR: Time series forecasting is extended here to longer-term prediction, obtained using the least-square support vector machines model, performed using the fast bootstrap methodology introduced in previous works.
Abstract: Time series forecasting is usually limited to one-step ahead prediction. This goal is extended here to longer-term prediction, obtained using the least-square support vector machines model. The influence of the model parameters is observed when the time horizon of the prediction is increased and for various prediction methods. The model selection to optimize the design parameters is performed using the fast bootstrap methodology introduced in previous works.
11 citations
01 Jan 2004
TL;DR: Amaury Lendasse, Damien Francois, Fabrice Rossi, Vincent Wertz, Michel Verleysen 1 Helsinki University of Technology Lab. Computer and Information Science, Neural Networks Research Centre, P.O.P. 5400, FIN-02015 HUT, Finlande, lendasse@hut.fi 2 Universite catholique de Louvain -machine learning group, CESAME, 4 av. G. Lemaitre, 1348 Lévain-la-Neuve, Belgium, francois@auto.ucl.ac.
Abstract: Amaury Lendasse, Damien Francois, Fabrice Rossi, Vincent Wertz, Michel Verleysen 1 Helsinki University of Technology – Lab. Computer and Information Science, Neural Networks Research Centre, P.O. Box 5400, FIN-02015 HUT, Finlande, lendasse@hut.fi 2 Universite catholique de Louvain – Machine Learning Group, CESAME, 4 av. G. Lemaitre, 1348 Louvain-la-Neuve, Belgique, francois@auto.ucl.ac.be 3 Projet AxIS, INRIA-Rocquencourt, Domaine de Voluceau, Rocquencourt, B.P. 105, 78153 Le Chesnay Cedex, France, Fabrice.Rossi@inria.fr. 4 Universite catholique de Louvain – Machine Learning Group, DICE, 3 place du Levant, 1348 Louvain-la-Neuve, Belgique, verleysen@dice.ucl.ac.be
2 citations
Posted Content•
TL;DR: In this article, an independent classification based on the style analysis, originally introduced by Sharpe in 1992 (Sharpe 1997), can be built, which leads to a classification of the funds based on characteristics extracted from their rates of returns and less prone to any manipulation.
Abstract: The ranking of investment funds regularly carried out by the financial press is a major concern for investors. In this context, there exists a serious risk of distortion between the announced strategy and the actual strategy implemented by the managers of the funds. Indeed these managers could try to have the funds they manage classified in a category that does not reflect the reality of their strategy, in order to be favourably compared to the performances of other funds in the same category. Our work shows how an independent classification based on the style analysis, originally introduced by Sharpe in 1992 (Sharpe 1997), can be built. This leads to a classification of the funds based on characteristics extracted from their rates of returns and less prone to any manipulation. The proposed classification is compared to a reference classification (ICDI and S&P). The analyses of the differences between the obtained results and, in particular, of their origin speak for themselves
2 citations
01 Jan 2004
TL;DR: The relationship between business plan quality and venture success is studied and a classification method using the information from the business plan is used in order to predict the success or the failure of each project.
Abstract: A business plan is a document presenting in a concise form the key elements describing a perceived business opportunity. It is used among others as a tool for evaluating the feasibility and profitability of a project. In this paper, we study the relationship between business plan quality and venture success. In particular, a classification method using the information from the business plan is used in order to predict the success or the failure of each project. The classification method presented in this paper is based on a linear piecewise approximation method known as Locally Pruned Lazy Learning Model.