Variable selection is one of the most fundamental problems in regression analysis. By sampling from the posterior distributions of candidate models, Bayesian variable selection via MCMC （Markov chain Monte-Carlo） is...Variable selection is one of the most fundamental problems in regression analysis. By sampling from the posterior distributions of candidate models, Bayesian variable selection via MCMC （Markov chain Monte-Carlo） is effective to overcome the computational burden of all-subset variable selection approaches. However, the convergence of the MCMC is often hard to determine and one is often not sure about if obtained samples are unbiased. This complication has limited the application of Bayesian variable selection in practice. Based on the idea of CFTP （coupling from the past）, perfect sampling schemes have been developed to obtain independent samples from the posterior distribution for a variety of problems. Here the authors propose an efficient and effective perfect sampling algorithm for Bayesian variable selection of linear regression models, which independently and identically sample from the posterior distribution of the model space and can efficiently handle thousands of variables. The effectiveness of the authors＇ algorithm is illustrated by three simulation studies, which have up to thousands of variables, the authors＇ method is further illustrated in SNPs （single nucleotide polymorphisms） association study among RA （rheumatoid arthritis） patients.展开更多
The aim of this research was to value, using a multiple regression model, the role of knowledge to guarantee the development in rural areas of European Union countries over 10 years. The main question was to find out ...The aim of this research was to value, using a multiple regression model, the role of knowledge to guarantee the development in rural areas of European Union countries over 10 years. The main question was to find out relationships among some variable, as the percentage of national Gross Domestic Product （GDP） used to improve the high training, and rural development in terms of agricultural labour units. The results underlined in 2001 as an high value of rural development, in terms of working force in agriculture, was identified in some countries of European Union characterised by a low value both in high training investments and also by a low value of Human Development Index, according to the definition of The Economist. The results in 2010 pointed out an inverse correlation among the dependent variable development in rural areas and the independent variables per capita GDP and national expenditure in advanced training, in percentage of national GDP. The learning by doing and by using, the introduction of advanced training in agriculture, using Long Life Learning measures of European Union, are important to improve the development of European rural areas but, sometimes, these actions are not perceived as something of useful.展开更多
A novel method based on machine learning is developed to estimate event-related potentials from single trial electroencephalography. This paper builds a basic framework using classification and an optimization model b...A novel method based on machine learning is developed to estimate event-related potentials from single trial electroencephalography. This paper builds a basic framework using classification and an optimization model based on this framework for estimating event-related potentials. Then the SingleTrialEM algorithm is derived by introducing a logistic regression model, which could be obtained by training before SingleTrialEM is used, to instantiate the optimization model. The simulation tests demonstrate that the proposed method is correct and solid. The advantage of this method is verified by the comparison between this method and the Woody filter in simulation tests. Also, the cognitive test results are consistent with the conclusions of cognitive science.展开更多
Mammalian carnivores are rarely considered for environmental reconstructions because they are extremely adaptable and their geographic range is usually large. However, the functional morphology of carnivore long bones...Mammalian carnivores are rarely considered for environmental reconstructions because they are extremely adaptable and their geographic range is usually large. However, the functional morphology of carnivore long bones can be indicative of locomotor behaviour as well as adaptation to specific kind of habitats. Here, different long bone ratios belonging to a subsample of extant large carnivores are used to infer palaeoecology of a comparative sample of Plio-Pleistocene fossils belonging to Italian paleo-communities. A multivariate long bone shape space reveals similarities between extant and fossil carnivores and multiple logistic regression models suggest that specific indices （the brachial and the Mt/F） can be applied to predict adaptations to grassland and tropical biomes. These functional indices exhibit also a phylogenetic signal to different degree. The brachial index is a significant predictor of adaptations to tropical biomes when phylogeny is taken into account, while Mt/F is not correlated anymore to habitat adaptations. However, the proportion of grassland-adapted carnivores in Italian paleo-communities exhibits a negative relationship with mean oxygen isotopic values, which are indicative of past climatic oscillations. As climate became more unstable during the Ice Ages, large carnivore guilds from the Italian peninsula were invaded by tropical/closed-adapted species. These species take advantage of the temperate forest cover that was more spread after 1.0 Ma than in the initial phase of the Quaternary （2.0 Ma） when the climate was more arid [Current Zoology 57 （3）： 269-283, 2011].展开更多
文摘Variable selection is one of the most fundamental problems in regression analysis. By sampling from the posterior distributions of candidate models, Bayesian variable selection via MCMC （Markov chain Monte-Carlo） is effective to overcome the computational burden of all-subset variable selection approaches. However, the convergence of the MCMC is often hard to determine and one is often not sure about if obtained samples are unbiased. This complication has limited the application of Bayesian variable selection in practice. Based on the idea of CFTP （coupling from the past）, perfect sampling schemes have been developed to obtain independent samples from the posterior distribution for a variety of problems. Here the authors propose an efficient and effective perfect sampling algorithm for Bayesian variable selection of linear regression models, which independently and identically sample from the posterior distribution of the model space and can efficiently handle thousands of variables. The effectiveness of the authors＇ algorithm is illustrated by three simulation studies, which have up to thousands of variables, the authors＇ method is further illustrated in SNPs （single nucleotide polymorphisms） association study among RA （rheumatoid arthritis） patients.
文摘The aim of this research was to value, using a multiple regression model, the role of knowledge to guarantee the development in rural areas of European Union countries over 10 years. The main question was to find out relationships among some variable, as the percentage of national Gross Domestic Product （GDP） used to improve the high training, and rural development in terms of agricultural labour units. The results underlined in 2001 as an high value of rural development, in terms of working force in agriculture, was identified in some countries of European Union characterised by a low value both in high training investments and also by a low value of Human Development Index, according to the definition of The Economist. The results in 2010 pointed out an inverse correlation among the dependent variable development in rural areas and the independent variables per capita GDP and national expenditure in advanced training, in percentage of national GDP. The learning by doing and by using, the introduction of advanced training in agriculture, using Long Life Learning measures of European Union, are important to improve the development of European rural areas but, sometimes, these actions are not perceived as something of useful.
基金supported by the National Natural Science Foundation of China (Grant No. 30670669)National Basic Research Program of China (Grant No. 2007CB947703)+1 种基金Natural Science Foundation of Fujian Province (Grant No. 2011J01344)Science and Technology Development Foundation of Fuzhou University (Grant No. 2009-XQ-25)
文摘A novel method based on machine learning is developed to estimate event-related potentials from single trial electroencephalography. This paper builds a basic framework using classification and an optimization model based on this framework for estimating event-related potentials. Then the SingleTrialEM algorithm is derived by introducing a logistic regression model, which could be obtained by training before SingleTrialEM is used, to instantiate the optimization model. The simulation tests demonstrate that the proposed method is correct and solid. The advantage of this method is verified by the comparison between this method and the Woody filter in simulation tests. Also, the cognitive test results are consistent with the conclusions of cognitive science.
文摘Mammalian carnivores are rarely considered for environmental reconstructions because they are extremely adaptable and their geographic range is usually large. However, the functional morphology of carnivore long bones can be indicative of locomotor behaviour as well as adaptation to specific kind of habitats. Here, different long bone ratios belonging to a subsample of extant large carnivores are used to infer palaeoecology of a comparative sample of Plio-Pleistocene fossils belonging to Italian paleo-communities. A multivariate long bone shape space reveals similarities between extant and fossil carnivores and multiple logistic regression models suggest that specific indices （the brachial and the Mt/F） can be applied to predict adaptations to grassland and tropical biomes. These functional indices exhibit also a phylogenetic signal to different degree. The brachial index is a significant predictor of adaptations to tropical biomes when phylogeny is taken into account, while Mt/F is not correlated anymore to habitat adaptations. However, the proportion of grassland-adapted carnivores in Italian paleo-communities exhibits a negative relationship with mean oxygen isotopic values, which are indicative of past climatic oscillations. As climate became more unstable during the Ice Ages, large carnivore guilds from the Italian peninsula were invaded by tropical/closed-adapted species. These species take advantage of the temperate forest cover that was more spread after 1.0 Ma than in the initial phase of the Quaternary （2.0 Ma） when the climate was more arid [Current Zoology 57 （3）： 269-283, 2011].
