上塘围钼多金属矿地质特征及成因探讨

矿床成因类型有两种:高温热液裂隙充填矿床和高温热液交代矿床,工业矿体为热液充填型石英矿脉,具有矿物品位较高的特点。根据矿区地球化学特征表明,Mo、W、Bi异常较好,新发现绿柱石(BeO),具有一定的找矿前景,下一步工作中应增加绿柱石矿的化验分析。     

传动角最优的测井仪推靠系统优化设计

通过对微球聚焦测井仪推靠系统主传动机构的运动学、动力学和行程传动角分析.按照实际工况要求,建立了以工作行程内推靠系统主传动机构传动角与期望传动角均方根波动最小为目标的多变量优化设计数学模型.并通过实例计算,对比分析优化前后测井仪推靠系统主传动机构工作行程内传动角的曲线,结果证明,优化效果显著,优化结果改善了推靠系统主传...     

求取水淹层混合地层水电阻率的方法探讨

目前,随着注水开发的深入,大部分油田逐步进入高含水中后期或特高含水阶段,相继进入二期或三期采油期,水淹情况日益严重,油田后期开采的难度也越来越大。水淹层地层水电阻率不仅是水淹层评价十分重要的基础参数,也是制约储层评价的难点。求取水淹层电阻率必须充分结合现阶段锰水开发条件,注意水性的变化及矿化度因注水加强而带来的影响,以达到更好的地质效果。本文针对尕斯油田已有相关资料,通过混合地层水电阻率图版的建立,从而为剩余油模型的建立提供依据。     

神经网络的知识获取与行为解释

神经网络的缺点之一在于它无法明显地表达其行为的含义.针对神经网络的不同结构,通过给出节点取值不同的约束条件,描述了如何从神经网络提取知识,并对神经网络的行为进行解释.举例说明了从神经网络提取产生式规划的策略.     

Visualization for Explanation of Deep Learning-Based Defect Detection Model Using Class Activation Map

Recently, convolutional neural network (CNN)-based visual inspection has been developed to detect defects on building surfaces automatically. The CNN model demonstrates remarkable accuracy in image data analysis; however, the predicted results have uncertainty in providing accurate information to users because of the “black box” problem in the deep learning model. Therefore, this study proposes a visual explanation method to overcome the uncertainty limitation of CNN-based defect identification. The visual representative gradient-weights class activation mapping (Grad-CAM) method is adopted to provide visually explainable information. A visualizing evaluation index is proposed to quantitatively analyze visual representations; this index reflects a rough estimate of the concordance rate between the visualized heat map and intended defects. In addition, an ablation study, adopting three-branch combinations with the VGG16, is implemented to identify performance variations by visualizing predicted results. Experiments reveal that the proposed model, combined with hybrid pooling, batch normalization, and multi-attention modules, achieves the best performance with an accuracy of 97.77%, corresponding to an improvement of 2.49% compared with the baseline model. Consequently, this study demonstrates that reliable results from an automatic defect classification model can be provided to an inspector through the visual representation of the predicted results using CNN models.     

两两比较模型的Why-not问题解释及排序

由于数据缺失，数据库用户通常无法获得查询结果中的预期答案.它被称为"Why-not问题"，即"为什么预期的元组不会出现在结果中".现有的方法通过列举可能的元组值来解释Why-not问题.枚举所给出解释的数量往往太大，无法由用户探索.完整性约束，如函数依赖，被用来排除不合格的解释.然而，许多属性在简化后解释中仅仅表示为变量，用户可能仍然无法理解.由于数据稀疏性，许多不合理的解释也会被推荐给用户.提出通过研究元组间两两比较关系，从而对Why-not问题的解释进行排序的方法.首先，重新定义为什么Why-not问题解释的形式没有变量，以便于用户理解；其次，对元组中的相等/不相等关系进行表示，提出在{0，1}表示的元组对的基础上学习统计模型，从而解决直接在原始数据上学习所带来的稀疏性问题，许多模型可以被用来推断概率，包括统计分布、分类和回归；最后，根据推断的概率对解释进行评价和排序.实验结果证明：利用统计、分类和回归方法计算两两关系概率分布的方法，可以为用户寻找Why-not问题的解释并返回较为高质量的解释.     

Novel insights gained from tagging walleye (Sander vitreus) with pop-off data storage tags and acoustic transmitters in Lake Ontario

Improvements in electronic tagging techniques provide new opportunities to gain insights into fish habitat selection and behaviours that have been difficult to capture using traditional assessment methods. However, data from acoustic telemetry studies in large freshwater systems may bias our understanding of fish habitat use and behaviour because of the typically low sampling frequency of transmitters as well as the limited spatial coverage and distribution of receivers in a waterbody. This study combined acoustic transmitters and pop-off data storage tags (pDSTs) on individual walleye in Lake Ontario to gain a better understanding of the feasibility and utility of double tagging a large nearshore freshwater fish. High frequency pDST data (every 2 s) revealed a novel diving behaviour by walleye which made repeated rapid dives beyond their standard daily depth range. Comparison of data from the two tag types showed that in this large freshwater system with low overall receiver coverage (∼4%), mean monthly and daily depth and temperature occupancy of walleye were similar, although many of the extreme values observed in the pDST data were not observed in the acoustic data. The accuracy of daily vertical distance travelled by walleye and summertime diving parameters was dependent on sampling frequency and only the pDST logging on a 2 s interval was able to provide reliable results. The results of this study show that novel insights can be gained, for fish large enough to handle the burden of multiples tags, ranging from spatial ecology to diving behaviours.     

Decision Boundary Visualization for Counterfactual Reasoning

Machine learning algorithms are widely applied to create powerful prediction models. With increasingly complex models, humans' ability to understand the decision function (that maps from a high-dimensional input space) is quickly exceeded. To explain a model's decisions, black-box methods have been proposed that provide either non-linear maps of the global topology of the decision boundary, or samples that allow approximating it locally. The former loses information about distances in input space, while the latter only provides statements about given samples, but lacks a focus on the underlying model for precise ‘What-If'-reasoning. In this paper, we integrate both approaches and propose an interactive exploration method using local linear maps of the decision space. We create the maps on high-dimensional hyperplanes—2D-slices of the high-dimensional parameter space—based on statistical and personal feature mutability and guided by feature importance. We complement the proposed workflow with established model inspection techniques to provide orientation and guidance. We demonstrate our approach on real-world datasets and illustrate that it allows identification of instance-based decision boundary structures and can answer multi-dimensional ‘What-If'-questions, thereby identifying counterfactual scenarios visually.     

基于现场总线的生产井录井仪及其专家系统

随着钻井生产由粗放钻井向精细钻井的发展和油气开发的决策需要，必须配备生产井录井仪，对生产井的钻井过程进行全过程监测和录井信息评价。针对生产井钻井过程的特点，设计研制了基于现场总线的小型化、配置灵活、功能开放，成本较低的油气田生产井录井仪及其专家系统。现场应用表明，研制开发的软硬件系统不仅能够为生产井的精细钻井和录井信息评价提供准确齐全的钻井参数，同时，设计开发的录井专家系统还能够实时准确地对钻井过程进行状态判断，实现了指导钻井的目的。