Knowing others' preferences degrades the quality of group decisions.

Results from 4 experiments demonstrate that learning the other group members' preferences at the beginning of a discussion impedes the solution of hidden profiles. In Experiments 1–3, participants who were not informed about their fellow group members' preferences were more likely to solve a hidden profile than those who received bogus information about the others' preferences. The negative effect of learning the others' preferences on decision quality was mediated by participants paying less attention to the information exchanged when they had been made aware of the others' preferences. Experiments 1 and 2 further ruled out that the effect of learning the others' preferences is due to participants bolstering their position or due to an increase in informational load. Experiment 3 showed that learning the other group members' preferences impedes the solution of hidden profiles even if one of the other members favors the correct alternative. Finally, Experiment 4 replicated these results in face-to-face interacting 3-person groups. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

基于OFDMA的分布式网络架构干扰分析及资源池概念

为了评估采用正交频分多址接入技术的多小区分布式网络架构的干扰状况，以广义分布式小区架构——群小区为例，具体分析了不采用功率控制、采用信干比平衡功率控制策略以及进一步采用逐步删除算法时的群小区架构小区间干扰状况，评估了与传统蜂窝架构相比较的系统中断性能．分析与仿真表明，相对于传统蜂窝结构，群小区架构的小区间干扰并没有大幅度的提高．并且由于群小区架构的系统资源是统一进行分配与管理的，系统性能较传统蜂窝架构有更大的提高．基于此分析，进一步提出了资源池概念．     

基于改进残差网络的扬尘图像识别方法

当前利用深度学习方法进行扬尘图像识别的研究较少,一些传统的方法使得扬尘图像的识别率较低.针对这种情况,提出一种基于改进残差网络的扬尘识别方法.该方法将ResNet-50网络应用到扬尘数据集中,并对其网络结构进行了改进.加入空间金字塔池化以解决输入图像尺寸不固定的问题,并且将金字塔池的策略改为平均池化,将扩大特征图的方法应用到主干网络中,有利于提取到更加细粒度的特征,提升模型的性能,从而提高识别率.实验结果表明,该方法具有很高的精确度,为扬尘识别提供了一种有效的方案.     

A bi-criteria model for the inventory aggregation problem under risk pooling

Inventory aggregation, also called Risk Pooling, is one of the most efficient ways to reduce the level of safety stocks thereby reducing inventory across the supply chain. Determining the best level of aggregation is a difficult problem and needs extensive study of all the possible scenarios that can affect this decision. Minimizing costs in a supply chain is no longer the sole priority of businesses. Maintaining a high level of responsiveness is also considered equally important. The conflicting nature of these two criteria makes the solution of the problem difficult. In this paper, we develop a bi-criteria nonlinear stochastic integer programming model to determine the best supply chain distribution network to meet customer demands, where minimizing costs while maintaining high levels of responsiveness is important. We develop a two-stage optimization algorithm to solve this problem.     

用于交通图像识别的改进尺度依赖池化模型

针对交通标志在自然场景中所占的比例较小、提取的特征量不足、识别准确率低的问题,提出改进的尺度依赖池化（SDP）模型用于小尺度交通图像的识别。首先,基于神经网络深卷积层具有较好的轮廓信息与类别特征,在SDP模型只提取浅卷积层特征信息的基础上,使用深卷积层特征补足型SDP（SD-SDP）映射输出,丰富特征信息;其次,因SDP算法中的单层空间金字塔池化损失边缘信息,使用多尺度滑窗池化（MSP）将特征池化到固定维度,增强小目标的边缘信息;最后,将改进的尺度依赖池化模型应用于交通标志的识别。实验结果表明,与原SDP算法比较,提取特征量增加,小尺度交通图像的识别准确率较好地提升。     

基于R CN N的问题相似度计算方法

在搜索引擎、问答系统中利用深度学习的方法计算问题相似度是NLP领域研究的热点.结合卷积神经网络(CNN)和长短记忆网络(LSTM),提出了递归卷积神经网络(RCNN)问句相似度的计算方法,首先利用双向递归神经网络提取上下文信息,然后采用1D卷积神经网络将词嵌入信息与上下文信息进行融合;再利用全局最大池化提取关键信息来完...     

Meta-Analytic Structural Equation Modeling: A Two-Stage Approach.

To synthesize studies that use structural equation modeling (SEM), researchers usually use Pearson correlations (univariate r), Fisher z scores (univariate z), or generalized least squares (GLS) to combine the correlation matrices. The pooled correlation matrix is then analyzed by the use of SEM. Questionable inferences may occur for these ad hoc procedures. A 2-stage structural equation modeling (TSSEM) method is proposed to incorporate meta-analytic techniques and SEM into a unified framework. Simulation results reveal that the univariate-r, univariate-z, and TSSEM methods perform well in testing the homogeneity of correlation matrices and estimating the pooled correlation matrix. When fitting SEM, only TSSEM works well. The GLS method performed poorly in small to medium samples. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

基于特征区域分割的立体视频视觉舒适度评价方法

为准确反映人眼在观看立体视频过程中的视觉舒适程度,提出了一种基于特征区域分割的立体视频视觉舒适度评价方法。首先自适应地分割提取出立体视频序列中的人眼的感兴趣区域,以及相邻帧间的运动区域,然后结合深度感知理论和空间合并技术,对视觉舒适度这一特征量进行建模,最后实现对立体视频视觉舒适度的评定。实验结果表明,使用本文的方法对立体视频视觉舒适度进行评价,可以很好的模拟人眼直接观测的视觉感受,对于视觉舒适度的研究具有重要的参考价值。     

融合可变形卷积网络的鱼眼图像目标检测

环视鱼眼图像具有目标形变大和图像失真的缺点,导致传统网络结构在对鱼眼图像进行目标检测时效果不佳.为解决环视鱼眼图像中由于目标几何畸变而导致的目标检测难度大的问题,提出一种基于可变形卷积网络的鱼眼图像目标检测方法.将Cascade_RCNN中固定的卷积层和池化层分别替换为可变形卷积层和可变形池化层,使用Resnet50网...