Improving reservoir volumetric estimations in petroleum resource assessment using discovery process models

The reservoir volumetric approach represents a widely accepted, but flawed method of petroleum play resource calculation. In this paper, we propose a combination of techniques that can improve the applicability and quality of the resource estimation. These techniques include： 1） the use of the Multivariate Discovery Process model （MDP） to derive unbiased distribution parameters of reservoir volumetric variables and to reveal correlations among the variables; 2） the use of the Geo-anchored method to estimate simultaneously the number of oil and gas pools in the same play; and 3） the crossvalidation of assessment results from different methods. These techniques are illustrated by using an example of crude oil and natural gas resource assessment of the Sverdrup Basin, Canadian Archipelago. The example shows that when direct volumetric measurements of the untested prospects are not available, the MDP model can help derive unbiased estimates of the distribution parameters by using information from the discovered oil and gas accumulations. It also shows that an estimation of the number of oil and gas accumulations and associated size ranges from a discovery process model can provide an alternative and efficient approach when inadequate geological data hinder the estimation. Cross-examination of assessment results derived using different methods allows one to focus on and analyze the causes for the major differences, thus providing a more reliable assessment outcome.     

Inverse problems

Positron emission tomography involves constructing an image of brain tissue from gamma rays counted at detectors surrounding the head. This is an inverse problem: how to measure a phenomenon from data taken from a derived distribution. We have noise and a loss of high frequency signal, both of which contribute to ill-conditioning. A brief and informal review of the mathematical and statistical methods available for handling inverse problems which are linear or linearizable is given.     

A New Nonparametric Pairwise Clustering Algorithm Based on Iterative Estimation of Distance Profiles

We present a novel pairwise clustering method. Given a proximity matrix of pairwise relations (i.e. pairwise similarity or dissimilarity estimates) between data points, our algorithm extracts the two most prominent clusters in the data set. The algorithm, which is completely nonparametric, iteratively employs a two-step transformation on the proximity matrix. The first step of the transformation represents each point by its relation to all other data points, and the second step re-estimates the pairwise distances using a statistically motivated proximity measure on these representations. Using this transformation, the algorithm iteratively partitions the data points, until it finally converges to two clusters. Although the algorithm is simple and intuitive, it generates a complex dynamics of the proximity matrices. Based on this bipartition procedure we devise a hierarchical clustering algorithm, which employs the basic bipartition algorithm in a straightforward divisive manner. The hierarchical clustering algorithm copes with the model validation problem using a general cross-validation approach, which may be combined with various hierarchical clustering methods.We further present an experimental study of this algorithm. We examine some of the algorithm's properties and performance on some synthetic and standard data sets. The experiments demonstrate the robustness of the algorithm and indicate that it generates a good clustering partition even when the data is noisy or corrupted.     

Leave One Out Error, Stability, and Generalization of Voting Combinations of Classifiers

We study the leave-one-out and generalization errors of voting combinations of learning machines. A special case considered is a variant of bagging. We analyze in detail combinations of kernel machines, such as support vector machines, and present theoretical estimates of their leave-one-out error. We also derive novel bounds on the stability of combinations of any classifiers. These bounds can be used to formally show that, for example, bagging increases the stability of unstable learning machines. We report experiments supporting the theoretical findings.     

Confirmatory factor analysis of the Wechsler Intelligence Scale for Children-Third Edition in an Australian clinical sample.

A confirmatory factor analysis was conducted on the Wechsler Intelligence Scale for Children-Third Edition (WISC-III; D. Wechsler, 1991) with a sample of 579 Australian children referred for assessment because of academic difficulties in the classroom. The children were administered the WISC-III as part of the initial eligibility determination process for funding of special education services. The children were aged between 6 years and 16 years 7 months. One-, two-, three-, and four-factor models were tested. The four-factor model proposed in the WISC-III manual fit the data significantly better than all other models tested. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

基于改进粒子群优化的SVM故障诊断方法

针对支持向量机(SVM)分类模型参数选取困难的问题,提出基于遗传免疫的改进粒子群优化算法,克服传统粒子群算法前期收敛快、后期易陷入局部最优的缺陷。将该算法与优化支持向量机分类模型相结合,建立基于遗传免疫粒子群和支持向量机的诊断模型,并用于轴承故障诊断中。结果表明,基于遗传免疫粒子群算法优化的SVM可实现对SVM分类模型参数的自动优化,并能提高SVM分类模型的故障诊断精度,对分散程度较大、聚类性较差的故障样本分类有较强的适用性。     

三维多尺度嵌套U结构CT影像肺结节检测

目的 肺结节检测在低剂量肺部计算机断层扫描(computed tomography, CT)筛查肺癌中具有重要意义。但由于结节大小、形状和密度的变化十分复杂,导致难以在低假阳性率下保证高的灵敏度,这限制了深度学习算法在常规临床实践中的肺结节自动诊断,建立具有良好结节检测性能的深度学习模型仍然是一个挑战。针对此问题,本文提出了一种基于3D ReSidual U(3D RSU)块的嵌套U结构的肺结节检测框架。方法 3D RSU块通过混合不同大小的感受场获得多尺度特征来丰富特征信息。而嵌套U结构允许网络获得更大分辨率的特征图,从而具有多层次深度特征,获得丰富的局部和全局信息,增强网络区分前景和背景的能力,进而提高微小结节等非显著性目标的检测性能。结果 该框架在公共肺结节(lung nodule analysis 16)挑战数据集上进行了评价。方法能够准确地检测出肺结节,灵敏度达到了97.2%,与基准方法相比,该方法灵敏度提高了2.6%,具有很高的灵敏度和特异性,在0.125、0.25、0.5、1、2、4、8共7个假阳率点的灵敏度平均值为86.4%,尤其是在每扫描0.25个假阳性上,灵敏度达到...     

基于交叉验证的BP算法的改进与实现

针对BP算法存在的收敛速度慢等问题提出改进方案,修改其相关参数并且提出如何选择合适的隐藏层节点个数.同时针对学习样本数据的有限性、BP算法易陷入局部最小值和容易出现过拟合等问题进行了研究,提出了采用多重交叉验证的再改进BP算法.仿真结果表明,交叉验证BP算法提高了网络学习的效率.     

利用改进的级联相关算法确定软件关键模块

利用交叉确认改进了级联相关算法,随后利用改进的级联相关算法,设计了多层前馈网络作为分类器,以软件模块的复杂性度量作为特征向量确定软件中的关键模块。最后以自行开发的维修性分配与预计(MAP)软件为例说明了采用改进的级联相关算法确定软件中的关键模块的优势。     

Use of data from long-term fertilizer experiments to model plant nitrogen uptake

Correct values for plant N-uptake are particularly important when correctly modelling the dynamics of nitrate leaching from agricultural soils. The general abundance of grain yield statistics represent a potential source of information for modelling plant N-uptake, provided that grain yield dry weight can be used to predict the total plant N-uptake (Vold & Søreng, 1997). In this paper, a nonlinear relationship for plant N-content as a function of grain yield dry weight and fertilizer N-level is derived. Data from long-term field experiments of grain dry weight (g N m^-2) and N-content in grains plus straw (g N m^-2) at harvest was used with nonlinear regression to estimate the parameters of the function. The parameters of a linear function were estimated similarly. Both models obtained could equally well describe the harvested N as a function of fertilizer N-level and grain dry weight, but the year-to-year variation of each fertilizer level was best described by the nonlinear model. An independent dataset, consisting of series of dry weight and N-content of grain yield from farm level experiments conducted in the same region (southeast Norway), was used to validate the nonlinear model.