The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Com o surgimento de uma grande quantidade de dados na ciência e na indústria, é urgente melhorar a precisão da previsão e reduzir a alta complexidade da regressão do processo gaussiano (GPR). No entanto, a aproximação global tradicional e a aproximação local têm deficiências correspondentes, tais como a aproximação global tende a ignorar as características locais, e a aproximação local tem o problema de sobreajuste. Para resolver esses problemas, é proposto um algoritmo de regressão de processo gaussiano em larga escala (RFFLT) combinando características aleatórias de Fourier (RFF) e aproximação local. 1) Para acelerar o tempo de treinamento, usamos os dados de entrada aleatórios do mapa de recursos de Fourier mapeados para o espaço de recursos aleatórios de baixa dimensão para processamento. A principal inovação do algoritmo é projetar recursos usando métodos de processamento linear rápido existentes, de modo que o produto interno dos dados transformados seja aproximadamente igual ao produto interno no espaço de recursos do kernel invariante de deslocamento especificado pelo usuário. 2) A máquina de comitê bayesiano robusta generalizada (GRBCM) baseada no método de informação mútua de Tsallis é usada na aproximação local, o que aumenta a flexibilidade do modelo e gera uma representação esparsa da distribuição de peso dos especialistas em comparação com trabalhos anteriores. O algoritmo RFFLT foi testado em seis conjuntos de dados reais, o que reduziu bastante o tempo de previsão da regressão e melhorou a precisão da previsão.
Hongli ZHANG
Yantai University
Jinglei LIU
Yantai University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copiar
Hongli ZHANG, Jinglei LIU, "Large-Scale Gaussian Process Regression Based on Random Fourier Features and Local Approximation with Tsallis Entropy" in IEICE TRANSACTIONS on Information,
vol. E106-D, no. 10, pp. 1747-1751, October 2023, doi: 10.1587/transinf.2023EDL8016.
Abstract: With the emergence of a large quantity of data in science and industry, it is urgent to improve the prediction accuracy and reduce the high complexity of Gaussian process regression (GPR). However, the traditional global approximation and local approximation have corresponding shortcomings, such as global approximation tends to ignore local features, and local approximation has the problem of over-fitting. In order to solve these problems, a large-scale Gaussian process regression algorithm (RFFLT) combining random Fourier features (RFF) and local approximation is proposed. 1) In order to speed up the training time, we use the random Fourier feature map input data mapped to the random low-dimensional feature space for processing. The main innovation of the algorithm is to design features by using existing fast linear processing methods, so that the inner product of the transformed data is approximately equal to the inner product in the feature space of the shift invariant kernel specified by the user. 2) The generalized robust Bayesian committee machine (GRBCM) based on Tsallis mutual information method is used in local approximation, which enhances the flexibility of the model and generates a sparse representation of the expert weight distribution compared with previous work. The algorithm RFFLT was tested on six real data sets, which greatly shortened the time of regression prediction and improved the prediction accuracy.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2023EDL8016/_p
Copiar
@ARTICLE{e106-d_10_1747,
author={Hongli ZHANG, Jinglei LIU, },
journal={IEICE TRANSACTIONS on Information},
title={Large-Scale Gaussian Process Regression Based on Random Fourier Features and Local Approximation with Tsallis Entropy},
year={2023},
volume={E106-D},
number={10},
pages={1747-1751},
abstract={With the emergence of a large quantity of data in science and industry, it is urgent to improve the prediction accuracy and reduce the high complexity of Gaussian process regression (GPR). However, the traditional global approximation and local approximation have corresponding shortcomings, such as global approximation tends to ignore local features, and local approximation has the problem of over-fitting. In order to solve these problems, a large-scale Gaussian process regression algorithm (RFFLT) combining random Fourier features (RFF) and local approximation is proposed. 1) In order to speed up the training time, we use the random Fourier feature map input data mapped to the random low-dimensional feature space for processing. The main innovation of the algorithm is to design features by using existing fast linear processing methods, so that the inner product of the transformed data is approximately equal to the inner product in the feature space of the shift invariant kernel specified by the user. 2) The generalized robust Bayesian committee machine (GRBCM) based on Tsallis mutual information method is used in local approximation, which enhances the flexibility of the model and generates a sparse representation of the expert weight distribution compared with previous work. The algorithm RFFLT was tested on six real data sets, which greatly shortened the time of regression prediction and improved the prediction accuracy.},
keywords={},
doi={10.1587/transinf.2023EDL8016},
ISSN={1745-1361},
month={October},}
Copiar
TY - JOUR
TI - Large-Scale Gaussian Process Regression Based on Random Fourier Features and Local Approximation with Tsallis Entropy
T2 - IEICE TRANSACTIONS on Information
SP - 1747
EP - 1751
AU - Hongli ZHANG
AU - Jinglei LIU
PY - 2023
DO - 10.1587/transinf.2023EDL8016
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E106-D
IS - 10
JA - IEICE TRANSACTIONS on Information
Y1 - October 2023
AB - With the emergence of a large quantity of data in science and industry, it is urgent to improve the prediction accuracy and reduce the high complexity of Gaussian process regression (GPR). However, the traditional global approximation and local approximation have corresponding shortcomings, such as global approximation tends to ignore local features, and local approximation has the problem of over-fitting. In order to solve these problems, a large-scale Gaussian process regression algorithm (RFFLT) combining random Fourier features (RFF) and local approximation is proposed. 1) In order to speed up the training time, we use the random Fourier feature map input data mapped to the random low-dimensional feature space for processing. The main innovation of the algorithm is to design features by using existing fast linear processing methods, so that the inner product of the transformed data is approximately equal to the inner product in the feature space of the shift invariant kernel specified by the user. 2) The generalized robust Bayesian committee machine (GRBCM) based on Tsallis mutual information method is used in local approximation, which enhances the flexibility of the model and generates a sparse representation of the expert weight distribution compared with previous work. The algorithm RFFLT was tested on six real data sets, which greatly shortened the time of regression prediction and improved the prediction accuracy.
ER -