面向不均衡数据集的过抽样算法

合成少数类过抽样技术（SMOTE）中的噪声样本可能参与合成新样本，所以难以保证新样本的合理性。针对这个问题，结合聚类算法提出了改进算法CSMOTE。该算法抛弃了SMOTE在最近邻间线性插值的思想，使用少数类的簇心与其对应簇中的样本进行线性插值合成新样本，并且对参与合成的样本进行了筛选，降低了噪声样本参与合成的可能。在六个实际数据集上，将CSMOTE算法与四个SMOTE的改进算法以及两种欠抽样算法进行了多次的对比实验，CSMOTE算法在所有数据集上均获得了最高的AUC值。实验结果表明，CSMOTE算法具有更高的分类性能，可以有效解决数据集中样本分布不均衡的问题。     

Pedestrian and Bicycle Volume Data Collection Using Drone Technology

ABSTRACT

Pedestrian and bicycle volume data is one of the most fundamental types of data for active transportation planning. However, they are not yet well developed. This paper explores an innovative method that uses drone technology to collect pedestrian and bicycle volume data. It confirms the feasibility of the technology as an alternative method to collect complex movements of pedestrians and bicycles. Furthermore, this paper presents a method that converts the video footage to a spatiotemporal dataset. The dataset includes not only the pedestrian and bicycle count data, but also their behavior and characteristics. The spatiotemporal data can become a valuable resource for a variety of active transportation planning practices and research, including the collection of pedestrian and bicycle volume data in parks and recreational areas, the study of collisions between pedestrians and bicycles, and the analysis of the social path. Since it is possible to envision that the development of technologies can overcome current technical difficulties, such as battery lifespan, it is worth considering the application of drone technology to active transportation planning practice and research.     

上海市遥感影像云服务平台设计与实现

针对上海市遥感影像管理和应用现状,采用镶嵌数据集为数据管理基本单元,云端影像处理服务为影像共享应用的基本手段,构建了上海市遥感影像云服务平台。解决了大规模遥感影像管理和应用存在的诸多问题,大幅缩短了影像应用的时间周期,并丰富了传统影像应用的手段,使遥感影像具有更强的适用性。     

跨模态脑图谱数据融合研究进展

脑图谱是研究脑结构和功能及脑疾病的基础,不同类型的脑图谱从不同角度提供了脑的组织模式或连接信息。随着图像采集和生物检测技术的发展,不同模态的脑影像和生物组学数据迅速增长。相较于单模态,多模态融合数据能够同时考察不同模态数据间的多元化信息,挖掘蕴含的未知新信息。因此,开展跨模态脑图谱数据融合研究有助于更全面地理解大脑的结构和功能,并辅助加深对脑发育、老化和病变机理的理解。本文根据参与融合的模态是否具有空间信息,将近年来有代表性的跨模态脑图谱融合技术分为脑影像融合和脑数据融合两大类。脑影像融合是指对宏观脑影像(磁共振等)和组织学脑影像(胞体染色、轴突染色等)等具有空间信息的数据进行融合,构建涵盖脑结构和功能信息的跨模态多尺度脑图谱,为研究宏观特征的介观机制以及介观特征的宏观表征提供了重要途径。脑数据融合是指对缺乏脑空间信息的生物大数据,包括基因组、电生理、认知和行为等,利用脑图谱提供精细空间信息,挖掘高维、异构生物大数据蕴含的信息,明确脑图谱的生理意义,并提升其应用价值。本文将针对这两类融合类型阐述国内外有代表性的研究进展,并对比国内外研究现状的差异。此外,为促进跨模态脑图谱数据融合领域的交流和发展,总结了部分有代表性的大样本公开数据集。最后讨论了当前该领域待解决的问题以及未来的发展趋势。     

Effectiveness of predicting tunneling-induced ground settlements using machine learning methods with small datasets

Prediction of tunneling-induced ground settlements is an essential task, particularly for tunneling in urban settings. Ground settlements should be limited within a tolerable threshold to avoid damages to aboveground structures. Machine learning (ML) methods are becoming popular in many fields, including tunneling and underground excavations, as a powerful learning and predicting technique. However, the available datasets collected from a tunneling project are usually small from the perspective of applying ML methods. Can ML algorithms effectively predict tunneling-induced ground settlements when the available datasets are small? In this study, seven ML methods are utilized to predict tunneling-induced ground settlement using 14 contributing factors measured before or during tunnel excavation. These methods include multiple linear regression (MLR), decision tree (DT), random forest (RF), gradient boosting (GB), support vector regression (SVR), back-propagation neural network (BPNN), and permutation importance-based BPNN (PI-BPNN) models. All methods except BPNN and PI-BPNN are shallow-structure ML methods. The effectiveness of these seven ML approaches on small datasets is evaluated using model accuracy and stability. The model accuracy is measured by the coefficient of determination (R²) of training and testing datasets, and the stability of a learning algorithm indicates robust predictive performance. Also, the quantile error (QE) criterion is introduced to assess model predictive performance considering underpredictions and overpredictions. Our study reveals that the RF algorithm outperforms all the other models with the highest model prediction accuracy (0.9) and stability (3.02 × 10^?27). Deep-structure ML models do not perform well for small datasets with relatively low model accuracy (0.59) and stability (5.76). The PI-BPNN architecture is proposed and designed for small datasets, showing better performance than typical BPNN. Six important contributing factors of ground settlements are identified, including tunnel depth, the distance between tunnel face and surface monitoring points (DTM), weighted average soil compressibility modulus (ACM), grouting pressure, penetrating rate and thrust force.     

黄土高原典型流域网格化逐时降水融合产品CMPA V1.0的精度分析

选取黄土高原无定河、窟野河和皇甫川流域为研究区,从水文应用角度评估了中国自动站与CMORPH融合的逐时降水产品(CMPA V1.0)的精度。结果表明,CMPA V1.0降雨数据总体上较雨量站实测值偏低;与6、9月相比,8月降水产品精度较低。随降水量级的增加,CMPA V1.0对降水事件发生频次的监测能力和对降水事件监测的稳定性有所降低,误报率整体升高;CMPA V1.0低估了小雨的发生频率及降水贡献率,高估了雨强大于2.5mm/h降雨的发生频率及降水贡献率;CMPA V1.0能合理地反映研究区降水的日变化过程,但低估了研究区的时段平均降水量和时段平均降水频率,高估了研究区14:00前的时段平均降水强度。     

Home energy efficiency and radon: An observational study

Exposure to radon gas is the second leading cause of lung cancer worldwide behind smoking. Changing the energy characteristics of a dwelling can influence both its thermal and ventilative properties, which can affect indoor air quality. This study uses radon measurements made in 470 689 UK homes between 1980 and 2015, linked to dwelling information contained within the Home Energy Efficiency Database (HEED). The linked dataset, the largest of its kind, was used to analyze the association of housing and energy performance characteristics with indoor radon concentrations in the UK. The findings show that energy efficiency measures that increase the airtightness of properties are observed to have an adverse association with indoor radon levels. Homes with double glazing installed had radon measurements with a significantly higher geometric mean, 67% (95% CI: 44, 89) greater than those without a recorded fabric retrofit. Those with loft insulation (47%, 95% CI: 26, 69) and wall insulation (32%, 95% CI: 11, 53) were also found to have higher radon readings. Improving the energy performance of the UK's housing stock is vital in meeting carbon emission reduction targets. However, compromising indoor air quality must be avoided through careful assessment and implementation practices.     

大规模连续中国手语数据集的创建与分析

绝大多数健听人不懂手语导致听障人在找工作、就医、法律咨询等各生活、工作领域中遇到了极大的沟通障碍，而手语翻译员需要提前预约，成本也非常高，所以很多科研工作者都开始利用机器学习来开发手语自动翻译器，但其中的大部分研究都因为受到了数据集规模和质量的影响而效果不佳。为解决上述矛盾和问题，创建了目前全球最大的中国连续手语数据集，并使用了考虑身体关节的位置、面部表情及手指关节的端到端的深度学习模型进行有效训练。结论突显了现代深度学习技术在识别复杂手语方面的巨大优势，针对较小子集的BLEU-4已达到30.8。     

A two-stage shape retrieval (TSR) method with global and local features

A robust two-stage shape retrieval (TSR) method is proposed to address the 2D shape retrieval problem. Most state-of-the-art shape retrieval methods are based on local features matching and ranking. Their retrieval performance is not robust since they may retrieve globally dissimilar shapes in high ranks. To overcome this challenge, we decompose the decision process into two stages. In the first irrelevant cluster filtering (ICF) stage, we consider both global and local features and use them to predict the relevance of gallery shapes with respect to the query. Irrelevant shapes are removed from the candidate shape set. After that, a local-features-based matching and ranking (LMR) method follows in the second stage. We apply the proposed TSR system to MPEG-7, Kimia99 and Tari1000 three datasets and show that it outperforms all other existing methods. The robust retrieval performance of the TSR system is demonstrated.