C4.5算法与粗糙集在数据挖掘中的研究

分类在数据挖掘中是一项非常重要的任务,而决策树方法和粗糙集方法则是分类的主要方法.在详细介绍决策树方法的经典算法C4.5算法和粗糙集方法的基础上,针对C4.5算法的不足,引人粗糙集方法,提出新算法IC4.5来弥补C4.5算法的缺陷.并设计出一个基于C4.5算法和粗糙集算法的分类器,最后通过实验来证明基于IC4.5算法的分类器在抗噪声方面要优于C4.5算法.     

Hydro-acoustic remote sensing of benthic biological communities on the shallow South East Australian continental shelf

Information regarding the composition and extent of benthic habitats on the South East Australian continental shelf is limited. In this habitat mapping study, multibeam echosounder (MBES) data are integrated with precisely geo-referenced video ground-truth data to quantify benthic biotic communities at Cape Nelson, Victoria, Australia. Using an automated decision tree classification approach, 5 representative biotic groups defined from video analysis were related to hydro-acoustically derived variables in the Cape Nelson survey area. Using a combination of multibeam bathymetry, backscatter and derivative products produced highest overall accuracy (87%) and kappa statistic (0.83). This study demonstrates that decision tree classifiers are capable of integrating variable data types for mapping distributions of benthic biological assemblages, which are important in maintaining biodiversity and other system services in the marine environment.     

基于LKH树和拉格朗日插值多项式的会议密钥分发方案

会议密钥的管理与分发是设计安全网络会议系统的关键问题之一。利用基于拉格朗日插值多项式的会议密钥分发方案,借鉴逻辑密钥层次树结构设计了1种会议密钥分发方案。在每次举行会议时,会议主席根据用户树计算参会者子树,利用参会者子树的根结点公钥构造拉格朗日插值多项式。减少了会议主席构造拉格朗日插值多项式的次数和计算开销。特别地,对构造的会议系统,当参会者的人数超过一定数目时,会议主席构造拉格朗日插值多项式的次数会随着参会者人数的增加而减少。另外,每个用户的私钥保持不变,避免了用户私钥的更新过程。     

基于点云强度的公园乔木检测方法

利用树干石灰涂层在激光点云中的高反射特性,本文提出一种基于点云强度的公园乔木检测方法。首先,使用三维激光扫描仪采集点云,并对点云进行去噪以及地面滤波。接着,对非地面点进行阈值分割,高强度部分包含了涂有石灰的树干。最后,对高强度点实施欧式聚类算法,通过尺寸与点数的约束,得到树干的聚类单元。通过实验证明,该方法具有自动化程度高、检测率高、误检率低等优势,满足公园调查的技术要求。     

基于多源数据的农业干旱监测模型构建

在全球气候变暖的大背景下,干旱事件发生越来越频繁,严重危害我国的粮食生产安全。构建准确的干旱监测模型不仅能够及时地反映出干旱事件的发生,同时可以为地方政府制定减灾保产措施提供科学支撑和保障。传统的气象干旱监测方法因为缺乏对植被本身需水状态和土壤供水信息的考虑旱情判定结果往往比实际情况偏重,而遥感监测指标大多只考虑了植被或土壤等单方面因素具有局限性,目前已有的综合干旱监测模型大多以气象指标为因变量,一方面需要数据资料较多参数复杂,另一方面模型准确度依赖于气象指标对当地农业干旱的响应能力,而同一气象指标在不同区域适应性存在差异,因此同样存在局限性。本文以河南省的冬小麦为研究对象,利用2001-2011年的EOS-MODIS数据产品以及气象站点监测数据,计算了标准化降水蒸散指数SPEI、植被状态指数VCI、温度状态指数TCI、温度植被状态指数TVDI,同时结合河南省农业气象灾害旬报对冬小麦受灾的记录,构建了基于决策树的定性农业干旱监测模型。测试集结果表明,模型综合了大气异常信息、植被状态信息以及土壤水分信息,优于单个指标的监测结果。另外,基于此模型监测了河南省2009年4-5月的干旱事件,结果与实情相符,能够较好地反映农业旱情的发生和空间演变情况。     

结合Gabor小波和形态学的高分辨率图像树冠提取方法

树冠信息的遥感提取能有效辅助森林参数反演、林分长势监测、树种识别等森林调查活动。随着遥感信息自动化提取的需求不断加强,本文基于高空间分辨率遥感数据,以滁州市皇甫山林场为研究区域,设计了一种结合Gabor小波和形态学的树冠提取方法。该方法首先采用Gabor小波提取出纹理特征,其次结合K-means聚类分析方法,对PCA降维后的纹理特征向量提取出阔叶林区,最后基于形态学理论降低影像噪声,并利用前景后景标记的分水岭方法进行单木树冠提取。经过与人工解译的树冠信息结果对比发现,在郁闭度较高的阔叶林区,该自动化方法提取树冠精度较高,分割准确率A_d为79.59%,F测度达到了79.00%能有效提供精确的单木树冠信息,为林业经济调查技术的发展具有一定的实践意义。     

How important is groundwater availability and stream perenniality to riparian and floodplain tree growth?

Riparian vegetation is important for stream functioning and as a major landscape feature. For many riparian plants, shallow groundwater is an important source of water, particularly in areas where rainfall is low, either annually or seasonally, and when extended dry conditions prevail for all or part of the year. The nature of tree water relationships is highly complex. Therefore, we used multiple lines of evidence to determine the water sources used by the dominant tree species Eucalyptus camaldulensis (river red gum), growing in riparian and floodplain areas with varying depth to groundwater and stream perenniality. Dendrometer bands were used to measure diel, seasonal, and annual patterns of tree water use and growth. Water stable isotopes (δ²H and δ¹⁸O) in plant xylem, soil water, and groundwater were measured to determine spatial and temporal patterns in plant water source use. Our results indicated riparian trees located on relatively shallow groundwater had greater growth rates, larger diel responses in stem diameter, and were less reactive to extended dry periods, than trees in areas of deep groundwater. These results were supported by isotope analysis that suggested all trees used groundwater when soil water stores were depleted at the end of the dry season, and this was most pronounced for trees with shallow groundwater. Trees may experience more frequent periods of water deficit stress and undergo reduced productivity in scenarios where water table accessibility is reduced, such as drawdown from groundwater pumping activities or periods of reduced rainfall recharge. The ability of trees to adapt to changing groundwater conditions may depend on the speed of change, the local hydrologic and soil conditions as well as the species involved. Our results suggest that E. camaldulesis growing at our study site is capable of utilizing groundwater even to depths >10 m, and stream perenniality is likely to be a useful indicator of riparian tree use of groundwater.     

Comparative analysis of predictive methods for drought durations: a case of monthly and annual streamflow droughts in Atlantic Canada

Methods based on the recursive probability, the extreme number theorem, and Markov chain (MC) concepts were applied to predict drought lengths (duration) on the standardized (termed as standardized hydrological index, SHI) sequences of monthly and annual river flows from Atlantic Canada. Results of the study indicated that the MC-based method is the most efficient, reliable and versatile method for predicting drought durations followed by the extreme-number-based method. The recursive-probability-based method was found to be computationally intensive and less efficient, although it provided a powerful means for calibrating the empirical plotting position formula needed in the MC-based method. The Weibull plotting position formula turned out to be a suitable measure of the exceedance probability in MC methodology for predicting drought lengths in Atlantic Canada. Based on results, it can be inferred that the MC-based method can be extended to MC2 and higher-order chains for predicting drought lengths on SHI sequences. The predictive capability of the extreme-number-theorem-based method is limited only to independent or weakly first-order persistent SHI sequences.

EDITOR D. Koutsoyiannis

ASSOCIATE EDITOR Q. Zhang     

Identifying the relationships between trophic states and their driving factors in the Shihmen Reservoir,Taiwan

Eutrophication has become a crucial issue for water resource management in recent years. In addition, reservoir trophic states are varied with environmental and water quality variables. The objectives of this study were to apply the DFA model to examine which water quality variables significantly affect variations of trophic state index (TSI) factors (i.e. total phosphorus (TP), chlorophyll-a (Chl-a), and Secchi disk transparency (SD)) and use classification and regression tree (CART) to determine the trophic states of the Shinmen Reservoir based on the levels of TSI factors during spring 2001–winter 2009. Results showed that the optimal DFA model contained one common trend (the underlying processes influencing trophic states, which can be rainfall intensity or runoff volume) and 7 explanatory variables. Turbidity (TB), pH, and dissolved oxygen (DO) influence concentrations of TP, while ammonium nitrogen (NH₃-N), organic nitrogen (O-N), and nitrate nitrogen (NO₃-N) control variations of Chl-a, and TB is related to SD. The CART model can specify trophic states only using two dominant driving factors, i.e. TP and Chl-a. The results of the CART illustrated that eutrophication could be occurred in the Shihmen Reservoir if TP is greater than 31.65 μg/L or if Chl-a is greater than 5.95 μg/L while TP concentration is less than 31.65 μg/L. Runoff nonpoint source pollution resulted from heavy storms may be the important factor affecting reservoir trophic states. Establishing vegetative filter strips along the riparian zone may able to effectively reduce this pollution in a reservoir. The integrated DFA and CART serves as good-fit relationships among trophic states, TSI factors, and water quality variables and provide control strategies for managing water quality in the Shihmen Reservoir.     

Probability-based seismic reliability assessment method for substation systems

Power systems play a key role in emergency rescue after an earthquake. Substations are among the most important components of a power system, so it is necessary to study their seismic reliability after an earthquake. Most studies, however, have focused on the seismic behavior of major electrical equipment or the reliability of the substation under ordinary operational conditions. The seismic reliability of substations as a system has not been thoroughly studied. This study proposes a new probability-based method to evaluate the seismic reliability of complex engineering systems such as substations. The proposed method developed the state tree to construct a greatly simplified system model that enables the failure probability to be calculated for the whole system, with explicit consideration of correlations among various components. A typical 220/110/10-kV substation was studied with this method, and the most critical components were identified.