ANN在森林资源预测中的应用研究

人工神经网络(ANN)方法是基于实例的方法,不需要考虑数学模型的内部结构,不需要假设前提条件,不需要人为地确定因子权重,作为一个黑箱综合地映射研究对象的整体性。应用人工神经网络多步预测方法对甘肃连城林场吐鲁沟营林区有林地面积进行预测,网络模型的最大相对误差为0 080 8％,最小相对误差达到0 0089％,平均为0.038 6％,表明预测值与实际值吻合程度很好,因此模型的精度较高,并且建模简单经预测,林场2000-2004年有林地面积稍有下降趋势,分别为2 873.2 hm2,2 618 7 hm2,2 484.5 hm2,2 346 hm2,2 171 6hm2。     

Including an additional systematic environmental effect within a generation in an evaluation model improves accuracy of prediction of breeding values in a closed herd of pigs

The present study evaluated the advantage of mixed‐model techniques over a selection index under different magnitudes of an additional systematic environmental effect (ASEE) in terms of accuracy of prediction and expected genetic gain. The data attempted to simulate a closed herd in a pig breeding program. The base population (G₀) consisted of 10 males and 50 females. Six generations (G₀ to G₅) were selected by using a selection index of three traits without overlapping. Additional systematic environmental constants with four levels in a generation were assigned from a uniform distribution at different ranges. Breeding values of animals in the last generation (G₅) were estimated on the basis of an index of individual phenotype (SI‐U), SI‐U adjusted for ASEE using a least‐squares mean (SI‐A), best linear unbiased prediction using an animal model excluding ASEE (AM‐E), and an animal model including ASEE (AM‐I). Accuracy of prediction and expected genetic gain were larger by the animal model than by the selection index, even if heritability of the traits selected was high and ASEE was set to zero. When ASEE was zero, the accuracy of prediction and expected genetic gain given by SI‐U and AM‐I were similar to those given by SI‐A and AM‐E, respectively. However, the differences in accuracy and expected gain between SI‐U and AI‐A and between AM‐I and AM‐E increased as the range of ASEE increased. It was concluded that selection based on an animal model was more effective than index selection, even if the herd environment was uniform and traits with high heritability were selected, and that it should be always included in an evaluation model, however slight any systematic environmental effect may be in a closed herd.     

Meta‐analysis of the energy and protein requirements of hair sheep raised in the tropical region of Brazil

The objective of this study was to estimate, through mathematical models, energy and protein requirements for maintenance and gain of hair sheep raised in the tropical region of Brazil. To determine the equation parameters, a meta‐analysis of seven independent experiments of nutrient requirements was performed, comprising a total of 243 experimental units (animals), which were conducted under tropical conditions, using hair sheep in growing and finishing phases and endowed of the following quantitative data for each animal: body weight (BW ), empty body weight (EBW ), average daily gain (ADG ), empty body gain (EBG ), heat production (HP ), metabolizable energy intake (MEI ), retained energy (RE ), metabolizable protein intake (MPI ) and body protein content. The regression equations generated were as follows: for Net Energy for maintenance, (NE _m): ; for Net Energy for gain, (NE _g): ; for Metabolizable Protein for maintenance,(MP _m): MPI _(g/day) = 24.8470 (±7.3646) + 560.28 (±99.6582) × EBG _(kg/day); for Net Protein for gain, (NP _g): . The NE _m requirement was 0.246 MJ EBW^?0.75 day^?1. The metabolizable energy for maintenance requirement was 0.391 MJ EBW^?0.75 day^?1. Considering an ADG of 100 g, the NE _g requirement ranged from 0.496 to 1.701 MJ/day for animals with BW ranging from 10 to 40 kg respectively. The efficiencies of use of the metabolizable energy for maintenance and gain were 0.63 and 0.36 respectively. The MP _m requirement was 3.097 g EBW^?0.75 day^?1. Considering an ADG of 100 g, the NP _g requirement ranged from 12.4 to 10.5 g/day for animals with BW ranging from 10 to 40 kg respectively. The total metabolizable energy and protein requirements were lower than those reported by the NRC and AFRC systems. Thus, our results support the hypothesis that nutrient requirements of hair sheep raised in tropical regions differ from wool sheep raised in temperate regions. Therefore, the use of the equations designed in this study is recommended.     

Geostatistical interpolation and aggregation of crop growth model outputs

Many crop growth models require daily meteorological data. Consequently, model simulations can be obtained only at a limited number of locations, i.e. at weather stations with long-term records of daily data. To estimate the potential crop production at country level, we present in this study a geostatistical approach for spatial interpolation and aggregation of crop growth model outputs. As case study, we interpolated, simulated and aggregated crop growth model outputs of sorghum and millet in West-Africa. We used crop growth model outputs to calibrate a linear regression model using environmental covariates as predictors. The spatial regression residuals were investigated for spatial correlation. The linear regression model and the spatial correlation of residuals together were used to predict theoretical crop yield at all locations using kriging with external drift. A spatial standard deviation comes along with this prediction, indicating the uncertainty of the prediction. In combination with land use data and country borders, we summed the crop yield predictions to determine an area total. With spatial stochastic simulation, we estimated the uncertainty of that total production potential as well as the spatial cumulative distribution function. We compared our results with the prevailing agro-ecological Climate Zones approach used for spatial aggregation. Linear regression could explain up to 70% of the spatial variation of the yield. In three out of four cases the regression residuals showed spatial correlation. The potential crop production per country according to the Climate Zones approach was in all countries and cases except one within the 95% prediction interval as obtained after yield aggregation. We concluded that the geostatistical approach can estimate a country’s crop production, including a quantification of uncertainty. In addition, we stress the importance of the use of geostatistics to create tools for crop modelling scientists to explore relationships between yields and spatial environmental variables and to assist policy makers with tangible results on yield gaps at multiple levels of spatial aggregation.     

基于“3S”技术的辽宁省某尾矿库扩容工程生态影响预测与评价

以辽宁省某尾矿库为例,基于GIS技术,利用2010年的TM遥感影像及地形图,预测该尾矿库扩容后的生态影响。分析结果显示:尾矿库扩容工程对生态的影响主要体现在土地利用、植被、生态景观和地形地貌的变化4个方面。     

Bioinformatics Analysis of Bone Morphogenetic Protein 4 (BMP4) of Sheep

The assay was aimed to explore the biological characteristics of bone morphogenetic protein 4 (BMP4) of sheep,NCBI,DNAMAN DNAStar,TMHMM Server v.2.0,PsortⅡ,SignalP various bioinformatical softwares were used to speculate the physical and chemical properties,hydrophobic property,phosphorylation site,conservative structure domain,protein secondary structure of BMP4 protein.Also,the three-dimensional structure was forecasted with the SWISS-MODEL Workspace software.The results indicated that the BMP4 of sheep had high homologies with the BMP4 of various species.The encoded protein was a hydrophilic protein which was unstable.There was no transmembrane regions and it was likely to be located in the nucleus.What was more,there was signal peptide and eighteen phosphorylation sites.Through the forecast of functional domains,the protein had two functional domains,including the transforming growth factor-beta (TGF-beta) superfamily and TGF-beta propeptide superfamily.The result was consistent with the function of BMP4 gene family,it also demonstrated that BMP4 was a growth factor and it had the function of signal transduction.The amino acid homology between the predicted 3D structure of protein and template 3bmp.1.A was 88.29%.The bioinformatics analysis of BMP4 gene could provide reference for the further study in practice.     

功能竹的定向培育及其可行性(英文)

功能竹是指经过人工优选、培育和生产出来的具有某种特殊作用或功能的栽培竹品种。功能竹的定向培育,就是根据市场前景广阔和价值最大化原则,运用有效的科学和技术手段并按照人类需求进行的功能竹培育。对于定向培育的功能竹,通常要求其性状相对一致、稳定,品质优异且便于推广,有利于组织科学化、标准化、产业化和规模化生产,目的是在同等时间和空间条件下,选择、培育和发展具有一个或多个指向明确的功能或特点的特色栽培竹品种,在不造成环境压力的情况下,尽可能满足消费者的物质和精神需求,并为生产者赢得更多的经济利益。作者通过认真分析论证,认为该思路不仅具有技术的可行性、资源的可行性和政策的可行性,而且具有经济的可行性和市场的可行性,这对立足中国自身资源特点发展绿色经济具有重要意义。     

基于贡献度随机森林模型的公司债信用风险实证分析

运用贡献度随机森林方法（CRF）方法探讨公司债财务指标比率与其违约率的关系.运用连续属性离散化方法（OB）进行财务指标最优降维；运用WOE变换进行模型变量约简.研究表明，CRF模型的分类性能显著优于其他模型，测试集评估总体正确率达90.47%，AUC统计量、AR比率及K-S值分别提升了2.6%、7.6%、4.38%，变量贡献度量化了各财务指标对违约率影响，为诠释随机森林预测机制提供了依据.     

基于Lasso和BP神经网络的组合预测及其应用——以居民消费支出预测为例

在变量选择的基础上，构建基于Lasso方法和BP神经网络的预测模型，并对我国城乡居民的消费支出进行预测，结果显示：基于Lasso方法和BP神经网络的组合预测精度要明显高于BP神经网络、Lasso方法的预测精度；在2014~2020年，我国农村居民消费增长率有所提升，城镇居民消费增长率减缓，城乡居民消费增长率之间的差距呈下降趋势，但短期内城乡居民消费差距依然难以缓和。     

花生黄曲霉生长预测模型的研究

通过对分离自花生的黄曲霉LYYSS03-3在不同温度(15-35℃)、不同相对湿度(水活度分别为0.5,0.7,0.9)下的菌落生长特性进行检测,采用Boltzmann和Logistic 模型对黄曲霉的生长曲线进行拟合,结果表明,在适宜温湿度阶段,花生黄曲霉的生长曲线更适合用Boltzmann模型拟合.