Distribution of CCR2-64I and SDF1-3 ' A alleles and HIV status in 7 ethnic populations of Cameroon

Limited information is available on the prevalence among rural Africans of host genetic polymorphisms conferring resistance to HIV-1 infection or slowing HIV disease progression.We report the allelic frequencies of the AIDS-related polymorphisms CCR2-64I, SDF1-3#A, and CCR5-D32 in 321 volunteers from 7 ethnic groups in Cameroon. Allelic frequencies differed among the 7 ethnic groups, ranging from 10.8% to 31.3% for CCR2-64I and 0.0% to 7.1% for SDF1-3#A. No CCR5-D32 alleles were found. HIV seroprevalence was 6.9% in the total population and peaked at younger ages in girls and women than in boys and men. Among 15- to 54-year-olds, HIV seroprevalence varied from 2.0% to 11.1% among the village populations. Conditional logistic regression analysis using data from boys and men aged 15 to 54 years showed the number of CCR2-64I alleles to be a significant risk factor for HIV seropositivity (odds ratio per allele adjusted for age and matched on ethnic group = 6.3, 95% confidence interval: 1.3–30.3); this association was not found in women. The findings are consistent with the hypothesis that CCR2-64I alleles may delay HIV disease progression without affecting susceptibility to infection among men. We did not observe this relation among women, and other factors, such as multiple pregnancies or maternal stressors (eg, breastfeeding), may have masked any protective effect of CCR2-64I alleles. Further study of this issue among women is warranted. SDF1-3#A did not differ between HIV-seropositive and HIV-seronegative individuals but wasassociated with increasing age among HIV-seronegative women, suggesting a protective effect against HIV-1 infection.

跨数据域迁移学习及相关问题的研究

在过去的二十年中，数据挖掘和机器学习受到了越来越多的关注。 这很大程度上是因为在互联网时代信息传播和积累的速度越来越快， 人工处理数据越来越困难，智能化及自动化的数据处理能力成为迫切的需求。 为此人们设计了很多学习算法，希望计算机能具有人类的学习能力，即只要训练一次，就 可以自动处理数据。 尽管这种学习能力已经在很多成功的应用中得到了验证，但它建立在一个重要的假设基础上，即训练数据与目标数据的一致性。 这意味着：根据训练数据得到的模型只适用于具有同样分布的目标数据。如果需要完成一个新的任务， 即使是与原任务非常相近的任务，原来训练好的模型也可能会失效。但是如果重新提供训练数据必将付出很高的成本。 因为两个任务之间存在的相似性，在新任务的训练过程中彻底丢弃原有的训练数据也是非常不合理的。 考虑到数据来源的差异性和训练数据的时效性在实际应用中普遍存在，有必要寻找更有效的解决途径。 迁移学习的提出正是为了解决上面的问题。传统的学习过程实际上是实现了从人到机器的知识迁移。 迁移学习则是研究从一个学习任务到另一个学习任务的知识迁移，以提高知识利用的效率。 这样的知识迁移将在缺乏训练数据和训练数据时效较短的情况下 大大降低学习的成本并提高学习的效率和自动化程度。 本文从跨数据域迁移学习入手，研究无监督迁移学习技术，以及在数据流环境下的有监督迁移学习技术， 在以下三个方面做出了创新性贡献： 在迁移学习中首次提出利用最大间隔方法在没有目标数据域的训练数据的情况下完成分类任务。 提出了两种算法，以迭代优化技术为基础，分别在函数层以及参数层实现了辅助任务到目标任务的知识迁移。 在多个公开的数据集中的实验表明，两种算法的分类准确率均优于现有的迁移学习算法。 在数据流分类任务中，针对概念漂移问题首次提出对概念漂移进行建模，来设计一种 可以自动适应数据分布变化的动态分类器。作为一种新的分类框架，可用于logistic regression和SVM等诸多分类模型。在实验中表明，所提出的算法有效避免了传统滑动 窗口方法导致的数据过拟合，实现了较高的分类准确率。 提出在具有多个节点的传感器网络中进行异常检测的新方法。利用主成分分析对数据空间进行变换，并根据能量阈值 对数据空间进行划分，构建异常子空间， 根据数据在异常子空间上的投影来检测异常数据点。基于数据点在异常子空间上的投影信息还可以进一步对异常来源 进行定位，并度量异常的大小。在实验中所提出的方法展现了较强的异常检测能力。

若尔盖湿地公路对两栖类的生态影响

本文考察了若尔盖高寒泥炭湿地公路对高原林蛙（Rana kukunoris）、倭蛙（Narorana pleskei）和岷山蟾蜍（Bufo minshanicus）的生态影响。分析了公路对两栖动物空间分布和栖息地利用的影响，并用IBM模型探讨其可能作用机制，考察了两栖动物公路死亡的季节差异及影响公路死亡空间分布的景观因素。最后通过对若尔盖高寒湿地两栖动物陆地核心栖息地的分析，为若尔盖路域栖息地的管理提供依据。 1. 对公路周边6个沼泽水凼群进行了调查，每个样地设置5条样线（距离公路10m、20m、50m、100m和150m）。调查表明，在繁殖季节(5月)，距离公路距离对高原林蛙和倭蛙的相对数量都有显著作用，其效应明显大于其他各项栖息地环境参数。公路导致高原林蛙和倭蛙在公路周边种群密度降低，其相对数量从距离公路100m处到公路边缘一直呈现逐渐降低的趋势。在繁殖季节，若尔盖高寒湿地的公路生态影响域大约在100-150m之间，这一距离远远大于森林栖息地中公路对两栖类的生态影响域（35-40 m）。 在繁殖后期（9月），对公路周边16个草地样点的样线调查表明，公路对周边高原林蛙和倭蛙密度分布并未造成显著影响。 2. 二次模型的拟合表明繁殖季节高原林蛙和倭蛙在公路周边的密度分布符合钟型曲线。前人对森林公路两侧两栖类分布的研究也显示了类似的规律。我们通过基于个体的模型，模拟在了公路边缘100单位距离内的栖息地空间，栖息地环境质量呈梯度变化，动物个体在其中通过随机运动寻找适宜的栖息地。拟合结果表明，动物个体仅仅依照简单的运动规则寻找适宜栖息地，这种活动就可以导致公路周边栖息地中的动物分布曲线出现3个局部峰。公路周边两栖动物的钟型分布曲线可能仅仅是个体寻找适宜栖息地过程中出现的临时性群体分布模式。 3. 在若尔盖高寒湿地，公路交通造成了大量两栖类死亡。但是公路两栖类动物死亡的季节分布很不均匀：5月、8月和9月死亡数量很高，而7月和10月死亡数量却很低。这种季节性差异和两栖类各个生活史阶段的迁移运动有密切的关系。利用景观参数的逻辑斯蒂回归模型显示，距离公路1000-2000m范围内的湿草地比例对三种两栖类公路死亡概率均有很强的贡献。湿草地这一栖息地类型分类中有大量的沼泽水体，是两栖类重要的繁殖点和取食点。两栖类公路死亡概率湿草地的关系从一个侧面表明，要维持一个区域较高的两栖类种群数量，需要1000-2000m半径范围内存在大面积的湿草地。 4. 高原林蛙和岷山蟾蜍不同性别和年龄个体分布点的水体距离存在显著差异。不同种类、年龄的两栖类分布点距离水体距离的差异可能是由于对水体的依赖性造成的。而相同种类、年龄段的个体中，高原林蛙雌性、岷山蟾蜍亚成体和雌性的体重与分布点距水体距离有显著负相关，这可能是因为体重更大的个体对水体的依赖性更弱。考虑到过大的陆地核心栖息地面积在实际保护工作中存在操作上的困难，因此我们认为可以以水体周边90%个体的分布区为低限确定3种两栖类的最小陆地核心栖息地。但是，在同样的水体距离-两栖类密度分布格局下，水体的面积和分形参数对最小陆地核心栖息地半径的确定有一定影响。 Ecological effects of alpine wetland road on Rana kukunoris, Narorana pleskei, Bufo minshanicus was studied in Zoige wetland. The effects of road on distribution of amphibians and its possible underline mechanism was discussed based on empirical data and computer simulation. Road killed amphibians was surveyed in different season and those landscape factor which could have impact on road killing distribution was analyses. Core terrestrial habitat of amphibians in Zoige wetland was discussed in the consideration of conservation management. 1. Six pool groups was investigated in breeding season (May) of R. kukunoris, N. pleskei. Five transects at distance of 10m, 20m, 50m, 100m and 150m from road edge was surveyed in each pool groups. There was a significant effects of distance from road edge on relative counts of R. kukunoris, N. pleskei, which is much important than effects of other environmental factors. Road caused the density of R. kukunoris, N. pleskei decreased from distance of 100m from road to 10m from road. Road ecological effect zone of alpine wetland for amphibians is about 100-150m. It is much wider than those of forest roads, which is about 35-40m. However, studies on 16 grassland near road showed no significant effect of road on amphibians after breeding season (Sep.). 2. Quadratic model fit indicated that the distribution of R. kukunoris and N. Pleskei followed a hump like curve. Previous studies on forest road showed similar results. A 100×100 habitat with gradual environment besides road was simulated with a individual-based model, and animal seek for suitable habitat with stochastic locomotion in it. Simulation results indicated that 3 density peak of animal distribution can emergent followed a simply rules. The hump like density cure could be a temporal swarm pattern during the process of individual seeking for habitat. 3. Road traffic caused mass death of amphibians in Zoige wetland. There was much road killed amphibians in May, Aug and Sep than those in July and Oct. The fluctuation of road kill could be related with migration of amphibians between seasons. Logistic regression of landscape variables indicated that wet grassland in 1000-2000m is essential to predict the probability of road kill. Wet grassland is an important breeding and forage habitat for amphibians. It also indicated that mass wet grassland in 1000-2000m is essential for maintain a big amphibian population. 4. There was significant differences among distance from aquatic site of subadults, female and males of R. kukunoris and B. Minshanicus. Possibly, it was because of their dependence on water. There was a significant negative relationship between distance from aquatic site and individuals body mass. Estimates of core habitat that are too large may make it difficult to establish protective regulations. The smallest suitable terrestrial core habitats were defined as the terrestrial habitats used during migration to and from the wetlands, and for foraging by 90% of any life stage (adults, and subadults) in a season. However, even with the same amphibian distribution pattern along the distance from aquatic sites, the radii of smallest suitable terrestrial core habitats will be varied with the fractal parameters of aquatic site.

滑坡泥石流信息遥感获取与空间预测

On the issue of geological hazard evaluation（GHE）, taking remote sensing and GIS systems as experimental environment, assisting with some programming development, this thesis combines multi-knowledges of geo-hazard mechanism, statistic learning, remote sensing (RS), high-spectral recognition, spatial analysis, digital photogrammetry as well as mineralogy, and selects geo-hazard samples from Hong Kong and Three Parallel River region as experimental data, to study two kinds of core questions of GHE, geo-hazard information acquiring and evaluation model. In the aspect of landslide information acquiring by RS, three detailed topics are presented, image enhance for visual interpretation, automatic recognition of landslide as well as quantitative mineral mapping. As to the evaluation model, the latest and powerful data mining method, support vector machine (SVM), is introduced to GHE field, and a serious of comparing experiments are carried out to verify its feasibility and efficiency. Furthermore, this paper proposes a method to forecast the distribution of landslides if rainfall in future is known baseing on historical rainfall and corresponding landslide susceptibility map. The details are as following: (a) Remote sensing image enhancing methods for geo-hazard visual interpretation. The effect of visual interpretation is determined by RS data and image enhancing method, for which the most effective and regular technique is image merge between high-spatial image and multi-spectral image, but there are few researches concerning the merging methods of geo-hazard recognition. By the comparing experimental of six mainstream merging methods and combination of different remote sensing data source, this thesis presents merits of each method ,and qualitatively analyzes the effect of spatial resolution, spectral resolution and time phase on merging image. (b) Automatic recognition of shallow landslide by RS image. The inventory of landslide is the base of landslide forecast and landslide study. If persistent collecting of landslide events, updating the geo-hazard inventory in time, and promoting prediction model incessantly, the accuracy of forecast would be boosted step by step. RS technique is a feasible method to obtain landslide information, which is determined by the feature of geo-hazard distribution. An automatic hierarchical approach is proposed to identify shallow landslides in vegetable region by the combination of multi-spectral RS imagery and DEM derivatives, and the experiment is also drilled to inspect its efficiency. (c) Hazard-causing factors obtaining. Accurate environmental factors are the key to analyze and predict the risk of regional geological hazard. As to predict huge debris flow, the main challenge is still to determine the startup material and its volume in debris flow source region. Exerting the merits of various RS technique, this thesis presents the methods to obtain two important hazard-causing factors, DEM and alteration mineral, and through spatial analysis, finds the relationship between hydrothermal clay alteration minerals and geo-hazards in the arid-hot valleys of Three Parallel Rivers region. (d) Applying support vector machine (SVM) to landslide susceptibility mapping. Introduce the latest and powerful statistical learning theory, SVM, to RGHE. SVM that proved an efficient statistic learning method can deal with two-class and one-class samples, with feature avoiding produce ‘pseudo’ samples. 55 years historical samples in a natural terrain of Hong Kong are used to assess this method, whose susceptibility maps obtained by one-class SVM and two-class SVM are compared to that obtained by logistic regression method. It can conclude that two-class SVM possesses better prediction efficiency than logistic regression and one-class SVM. However, one-class SVM, only requires failed cases, has an advantage over the other two methods as only "failed" case information is usually available in landslide susceptibility mapping. (e) Predicting the distribution of rainfall-induced landslides by time-series analysis. Rainfall is the most dominating factor to bring in landslides. More than 90% losing and casualty by landslides is introduced by rainfall, so predicting landslide sites under certain rainfall is an important geological evaluating issue. With full considering the contribution of stable factors (landslide susceptibility map) and dynamic factors (rainfall), the time-series linear regression analysis between rainfall and landslide risk mapis presented, and experiments based on true samples prove that this method is perfect in natural region of Hong Kong. The following 4 practicable or original findings are obtained: 1) The RS ways to enhance geo-hazards image, automatic recognize shallow landslides, obtain DEM and mineral are studied, and the detailed operating steps are given through examples. The conclusion is practical strongly. 2) The explorative researching about relationship between geo-hazards and alteration mineral in arid-hot valley of Jinshajiang river is presented. Based on standard USGS mineral spectrum, the distribution of hydrothermal alteration mineral is mapped by SAM method. Through statistic analysis between debris flows and hazard-causing factors, the strong correlation between debris flows and clay minerals is found and validated. 3) Applying SVM theory (especially one-class SVM theory) to the landslide susceptibility mapping and system evaluation for its performance is also carried out, which proves that advantages of SVM in this field. 4) Establishing time-serial prediction method for rainfall induced landslide distribution. In a natural study area, the distribution of landslides induced by a storm is predicted successfully under a real maximum 24h rainfall based on the regression between 4 historical storms and corresponding landslides.

中老年危机热线来电者的应激源与自杀危险性

Objective: Psychosocial crisis and psychiatric disorders are two stressors for suicide action. This study will explore the differences on demographic characteristics, severity of depression, and suicidality of middle-aged and elder crisis line callers under the influences of psychosocial crisis or psychiatric disorders or two simultaneously-mixed stressors, in order to develop effective intervention strategies for crisis line. Methods: Analysis data of 1,092 cases selected from national crisis line callers aged 45 and over who were assessed with “Suicide risk assessment” during the period from December, 2002 to December, 2008. The sample were divided into three groups of psychosocial crisis, mental health problems, and mixed-stressors of three types of general callers (48.2%, 32.3%, 19.5%), callers with current suicide ideation (43.7%, 33.0%, 23.3%) and callers attempted suicide 2 weeks prior to the call (33.6%, 42.3%, 24.1%) respectively according to the operators’ judgments of the callers’ claimed difficult situations and classification system of crisis line database. X2 test and Tukey-type and Multinomial Logistic Regression multiple comparison methods are applied to analysis the differences of the three groups. Results: In agreement with previous studies, more females (71.3%, X2=13.45, P<0.001), especially females influenced by relationship stressors (76.8%, X2=25.12, P<0.001) made the call for crisis. Among general callers, the check-out rates of Major Depression Episode of mixed-stressor callers (78.5%, P<0.001) and problem callers (68.7%, P<0.05) were significantly higher than that of crisis callers (57.1%). The check-out rates of suicide ideation of mixed-stressor callers (71.4%) were significantly higher than that in crisis callers (53.8%, P<0.001) and problem callers (60.9%, P<0.05). The check-out rates of prior suicide attempts of mixed-stressor (16.6%, P<0.05) and problem callers (18.5%, P<0.01) were significantly higher than that of crisis callers (9.8%). More than half of the mixed-stressor callers (51.8%) reported over 50% degree of hopelessness, which was significantly higher than that of crisis callers (35.6%, P<0.01) and problem callers (38.2%, P<0.05). Fewer crisis callers sought medical help than problem and mixed-stressor callers among three types of callers (X2=241.35, 146.56, 50.87; P<0.001). Compare to non-compound crisis callers, the proportion of minor, severe depression and prior depression diagnosis (14.0% vs. 17.4%; 54.9% vs. 65.2%; 0 vs. 2.2%; X2=14.35，P<0.01), suicide ideation (51.1% vs. 64.0%, P<0.05) and prior suicide attempts (8.4% vs. 15.0%, P<0.05) in compound crisis callers were significantly higher. There were more compound crisis callers with over 50% hopelessness (51.9% vs. 31.0%，X2=11.96，P<0.01). Conclusion: As predicted, among middle-aged and elderly participants, mixed-stressor and compound crisis callers were higher in degree of severity of depression and suicidality. Intervention strategies should be developed addressing to specific stressor or stressors. The promotions of crisis callers’ medical help seeking behavior need to be emphasized.