ID 原文 译文
58248 由于潜在高价值旅客当前乘机历史记录少，较难被航空公司准确发现并关注. Potential high-value passengers can not be effectively discovered by airways due to the limitedhistorical booking records of passengers.
58249 对此，提出基于出行意图的潜在高价值旅客发现概率模型. Aiming at this problem，a probabilistic model for discovering potential high-value passengers based on trip purposes mining is proposed.
58250 首先建立一个基于统计的潜在高价值旅客发现概率模型，再将旅客出行意图引入概率模型，发现旅客潜在航线需求，优化旅客潜在价值计算，从而通过出行意图发现潜在高价值旅客. Firstly，we present a probabilistic model based on statistics to measure the value of passengers. Then，trip purposes are introduced intothe model to discover potential airline demands of each passenger and to optimize passenger potential value calculation. Therefore，potential high-value passengers can be discovered through the trip purposesmining.
58251 实验结果表明，相比于次数法、里程法以及 ＲFM 模型等传统的旅客价值度量方法，基于出行意图的潜在高价值旅客发现概率模型能够有效识别潜在高价值旅客. Experiments show that the proposed model can identify the potential high-value passengers moreaccurately than the traditional passenger value evaluation methods based on the passengers’cumulativenumber of flight times，passengers’cumulative mileage and recency frequency monetry model.
58252 经典物流配送模型的目标、约束条件不够全面，在实际应用中存在一定缺陷，对此，构建了时间窗和油耗取送一体化的物流配送路径优化模型( PDVＲPTF) . Aiming at the problem that the classic logistics distribution model considers the target，theconstraints are not comprehensive enough and there are certain defects in the practical application，a integrated pickup and distribution vehicle routing problem on the basis of the classical model considering timewindow and fuel consumption ( PDVＲPTF) is constructed.
58253 设计了一种基于 k-medoids 动态聚类混合拓扑结构粒子群算法，解决了经典粒子群算法在求解此类模型时容易陷入局部最优解的问题. Hybrid topological structure of particle swarmoptimization based on k-medoids dynamic clustering is designed，which solves the problem that classicalparticle swarm optimization is easy to fall into local optimal solution when solving such models.
58254 仿真结果表明，改进型粒子群算法能很好地跳出局部最优解，并快速收敛于全局最优解，且该算法可有效求解物流配送路径优化的问题. The simulation results show that the improved particle swarm optimization can jump out of the local optimal solutionquickly and converge to the global optimal solution quickly，which solve the logistics distribution path optimization problem effectively.
58255 为了解决图像情感分析中存在的情感鸿沟和大的类内方差问题，提出了一种可以同时利用视觉模态和文本模态之间的深度潜在关联、视觉模态的深度线性判别和图像中层语义融合的弱监督方法. In order to alleviate affective gap and large intra-class variance existing in visual sentimentanalysis，firstly a new method is proposed，which exploits simultaneously not only deep latent correlationsbetween visual and textual modalities，but also deep linear discrimination of visual modality and weak supervision of mid-level semantic features of images.
58256 利用多模态深度网络结构找到一个视觉模态和文本模态之间最大深度关联且视觉模态具有深度判别性的潜在嵌入空间，并在该潜在空间中将文本的语义映射特征迁移到图像的判别性视觉映射特征中; The method uses multimodal deep network architectureto find a latent embedding space in which deep correlations between visual and textual modalities aremaximized，and at the same time there is a deep discrimination on visual modality. In the latent space，the extracted semantic feature of texts can be transferred to the extracted discriminant visual feature of images.
58257 结合注意力机制，设计涵盖潜在空间中映射特征的注意力网络，用于情感分类. Secondly based on the usfulness of attention mechanism，an attention network is presented，whichaccepts the extracted features in the latent space as input and is trained as a sentiment classifier.