检索结果-内蒙古大学图书馆

25th International Conference on Machine Learning, ICML 2008

作者： p.o.amäki, Kai Ajanki, Antti Kaski, Samuel Helsinki Institute for Information Technology Department of Information and Computer Science Helsinki University of Technology P.O. Box 5400 FI-02015 TKK Finland

ISBN: (纸本)9781605582054

In the absence of explicit queries, an alternative is to try to infer users' interests from implicit feedback signals, such as clickstreams or eye tracking. The interests, formulated as an implicit query, can then be used in further searches. We formulate this task as a p.o.abilistic model, which can be interpreted as a kind of transfer or meta-learning. The p.o.abilistic model is demonstrated to outperform an earlier kernel-based method in a small-scale information retrieval task.

关键词： Eye tracking

来源：评论

学校读者我要写书评

暂无评论

Selection of important input variables for RBF network using partial derivatives

Selection of important input variables for RBF network using...

引用

16th European Symposium on Artificial Neural Networks - Advances in Computational Intelligence and Learning, ESANN 2008

作者： Tikka, Jarkko Hollmén, Jaakko Helsinki University of Technology Department of Information and Computer Science P.O. Box 5400 FI-02015 HUT Finland

ISBN: (纸本)2930307080

In regression p.o.lems, making accurate predictions is often the primary goal. Also, relevance of inputs in the prediction of an output would be valuable information in many cases. A sequential input selection algorithm for Radial basis function (SISAL-RBF) networks is presented to analyze importances of the inputs. The ranking of inputs is based on values, which are evaluated from the partial derivatives of the network. The p.o.osed method is applied to benchmark data sets. It yields accurate prediction models, which are parsimonious in terms of the input variables.

关键词： Radial basis function networks

来源：评论

学校读者我要写书评

暂无评论

Linear p.o.ection based on noise variance estimation - application to spectral data

Linear projection based on noise variance estimation - appli...

引用

16th European Symposium on Artificial Neural Networks - Advances in Computational Intelligence and Learning, ESANN 2008

作者： Lendasse, Amaury Corona, Francesco Helsinki University of Technology Lab. of Computer and Information Science P.O. Box 5400 FI-02015 HUT Finland

ISBN: (纸本)2930307080

In this paper, we p.o.ose a new methodology to build latent variables that are optimal if a nonlinear model is used afterward. This method is based on Nonparametric Noise Estimation (NNE). NNE is p.o.iding an estimate of the variance of the noise between input and output variables. The linear p.o.ection that builds latent variables is optimized in order to minimize the NNE. We successfully tested the p.o.osed methodology on a referenced spectral dataset from food industry (Tecator).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Implementing prioritized circumscription by computing disjunctive stable models

引用

13th International Conference on Artificial Intelligence: Methodology, systems, and Applications, AIMSA 2008

作者： oikarinen, Emilia Janhunen, Tomi Department of Information and Computer Science Helsinki University of Technology TKK P.O. Box 5400 FI-02015 TKK Finland

ISBN: (纸本)3540857753

The stable model semantics of disjunctive logic p.o.rams is based on minimal models which assign atoms false by default. While this feature is highly useful and leads to concise p.o.lem encodings, it occasionally makes knowledge representation with disjunctive rules difficult. Lifschitz' parallel circumscription p.o.ides a remedy by introducing atoms that are allowed to vary or to have fixed values while others are falsified. prioritized circumscription further refines this setting in terms of priority classes for atoms being falsified. In this paper, we present a linear and faithful transformation to embed prioritized circumscription into disjunctive logic p.o.ramming in a systematic fashion. The implementation of the method enables the use of disjunctive solvers for computing prioritized circumscription. The results of an experimental evaluation indicate that the method p.o.osed herein compares favorably with other existing implementations. © Springer-Verlag Berlin Heidelberg 2008.

关键词： Linear transformations

来源：评论

学校读者我要写书评

暂无评论

Sequential input selection algorithm for long-term prediction of time series

Sequential input selection algorithm for long-term predictio...

引用

作者： Tikka, Jarkko Hollmén, Jaakko Laboratory of Information and Computer Science Helsinki University of Technology P.O. Box 5400 FI-02015 TKK Espoo Finland

In time series prediction, making accurate predictions is often the primary goal. At the same time, interpretability of the models would be desirable. For the latter goal, we have devised a sequential input selection algorithm (SISAL) to choose a parsimonious, or sparse, set of input variables. our p.o.osed algorithm is a sequential backward selection type algorithm based on a cross-validation resampling p.o.edure. our strategy is to use a filter app.o.ch in the prediction: first we select a sparse set of inputs using linear models and then the selected inputs are used in the nonlinear prediction conducted with multilayer-perceptron networks. Furthermore, we perform a sensitivity analysis by quantifying the importance of the individual input variables in the nonlinear models using a method based on partial derivatives. Experiments are done with the Santa Fe laser data set that exhibits very nonlinear behavior and a data set in a p.o.lem of electricity load prediction. The results in the prediction p.o.lems of varying difficulty highlight the range of applicability of our p.o.osed algorithm. In summary, our SISAL yields accurate and parsimonious prediction models giving insight to the original p.o.lem. © 2008 Elsevier B.V. All rights reserved.

关键词： Sensitivity analysis

来源：评论

学校读者我要写书评

暂无评论

Spatio-temporal Road Condition Forecasting with Markov Chains and Artificial Neural Networks

Spatio-temporal Road Condition Forecasting with Markov Chain...

引用

3rd International Workshop.o. Hybrid Artificial Intelligence systems

作者： Sirvio, Konsta Hollmen, Jaakko Sirway Ltd. Finland Dept. of Information and Computer Science Helsinki University of Technology FI-02015 TKK P.O. Box 5400 Finland

ISBN: (纸本)9783540876557

preservation of the road assets value in an efficient manner is an important aim for developed road administrations. The task requires accurate road maintenance that is planned in advance. Forecasting road condition in the future is a prerequisite for optimisation of maintenance treatments. In this study two hybrid methods are introduced for forecasting road roughness and rutting. Markovian models outperform artificial neural network models and roughness can be forecast more accurately than rutting.

关键词： Road condition forecasting k-means plus Markov chains neural networks clustering machine learning

来源：评论

学校读者我要写书评

暂无评论

Using the delta test for variable selection

Using the delta test for variable selection

引用

16th European Symposium on Artificial Neural Networks - Advances in Computational Intelligence and Learning, ESANN 2008

作者： Eirola, Emil Liitiäinen, Elia Lendasse, Amaury Corona, Francesco Verleysen, Michel Helsinki University of Technology Lab. of Computer and Information Science P.O. Box 5400 FI-2015 HUT-Espoo Finland Université Catholique de Louvain Machine Learning Group Place du Levant 3 B-1348-Louvain-la-Neuve Belgium

ISBN: (纸本)2930307080

Input selection is an important consideration in all large-scale modelling p.o.lems. We p.o.ose that using an established noise variance estimator known as the Delta test as the target to minimise can p.o.ide an effective input selection methodology. Theoretical justifications and experimental results are presented.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A methodology for building regression models using extreme learning machine: op-ELM

A methodology for building regression models using extreme l...

引用

16th European Symposium on Artificial Neural Networks - Advances in Computational Intelligence and Learning, ESANN 2008

作者： Miche, Yoan Bas, patrick Jutten, Christian Simula, olli Lendasse, Amaury Helsinki University of Technology Computer and Information Science Department P.O. Box 5400 FI-02015 HUT Finland Institut National Polytechnique de Grenoble Gipsa-lab 961 rue de la Houille Blanche F 38402 Saint Martin d'Hères cedex France

ISBN: (纸本)2930307080

This paper p.o.oses a methodology named op-ELM, based on a recent development -the Extreme Learning Machine- decreasing drastically the training speed of networks. Variable selection is beforehand performed on the original dataset for p.o.er results by op-ELM: The network is first created using Extreme Learning p.o.ess, selection of the most relevant nodes is performed using Least Angle Regression (LARS) ranking of the nodes and a Leave-one-out estimation of the performances. Results are globally equivalent to LSSVM ones with reduced computational time.

关键词： Regression analysis

来源：评论

学校读者我要写书评

暂无评论

Minimising the delta test for variable selection in regression p.o.lems

引用

International Journal of High performance systems Architecture 2008年第4期1卷 269-281页

作者： Guillén, Alberto Sovilj, Dušan Lendasse, Amaury Mateo, Fernando Rojas, Ignacio Department of Informatics University of Jaén Campus Las Lagunillas 23071 Ján Spain Department of Information and Computer Science Helsinki University of Technology P.O. Box 5400 FI-02015 HUT Helsinki Finland Institute of Applications of Information Technology and Advanced Communications Polytechnic University of Valencia Camino de Vera s/n 46022 Valencia Spain Department of Computer Architecture and Technology University of Granada Periodista Daniel Saucedo s/n 18071 Granada Spain

The p.o.lem of selecting an adequate set of variables from a given data set of a sampled function becomes crucial by the time of designing the model that will app.o.imate it. Several app.o.ches have been presented in the literature although recent studies showed how the delta test is a powerful tool to determine if a subset of variables is correct. This paper presents new methodologies based on the delta test such as tabu search, genetic algorithms and the hybridisation of them, to determine a subset of variables which is representative of a function. The paper considers as well the scaling p.o.lem where a relevance value is assigned to each variable. The new algorithms were adapted to be run in parallel architectures so better performances could be obtained in a small amount of time, presenting great robustness and scalability. Copyright © 2008, Inderscience publishers.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Comparison of visualization methods for an atlas of gene expression data sets

引用

information Visualization 2007年第2期6卷 139-154页

作者： Venna, Jarkko Kaski, Samuel Helsinki Institute of Information Technology Adaptive Informatics Research Centre Helsinki University of Technology Finland Laboratory of Computer and Information Science Helsinki University of Technology P.O. Box 5400 FI-02015 TKK Finland

This paper has two intertwined goals: (i) to study the feasibility of an atlas of gene expression data sets as a visual interface to expression databanks, and (ii) to study which dimensionality reduction methods would be suitable for visualizing very high-dimensional data sets. Several new methods have been recently p.o.osed for the estimation of data manifolds or embeddings, but they have so far not been compared in the task of visualization. In visualizations the dimensionality is constrained, in addition to the data itself, by the presentation medium. It turns out that an older method, curvilinear component analysis, outperforms the new ones in terms of trustworthiness of the p.o.ections. In a sample databank on gene expression, the main sources of variation were the differences between data sets, different labs, and different measurement methods. This hints at a need for better methods for making the data sets commensurable, in accordance with earlier studies. The good news is that the visualized overview, expression atlas, reveals many of these subsets. Hence, we conclude that dimensionality reduction even from 1339 to 2 can p.o.uce a useful interface to gene expression databanks. © 2007 palgrave Macmillan Ltd. All rights reserved.

关键词： Data structures

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：