在基于位置的社交网络中,兴趣点实时推荐数据和用户签到数据存在高稀疏性问题。提出一种基于时间效应的混合推荐模型。通过用户潜在兴趣点数据模型计算用户时间行为影响分数和地理位置影响分数,并用线性统一模型进行处理,选取Top S 个...在基于位置的社交网络中,兴趣点实时推荐数据和用户签到数据存在高稀疏性问题。提出一种基于时间效应的混合推荐模型。通过用户潜在兴趣点数据模型计算用户时间行为影响分数和地理位置影响分数,并用线性统一模型进行处理,选取Top S 个兴趣点作为用户的潜在兴趣点。将用户的潜在签到记录引入基于时间效应的矩阵分解模型中,考虑时间差异性和连续性对推荐结果的影响,在此基础上进行优化求解,提出推荐策略。实验结果表明,与LRT模型、UTE+SE模型相比,该模型的推荐效果较好,其准确率和召回率最高可达0.103 4和0.111 8。展开更多
Multi-channel Global Positioning System (GPS) satellite signal simulator is used to provide realistic test signals for GPS receivers and navigation systems. In this paper, signals arriving the antenna of GPS receiver ...Multi-channel Global Positioning System (GPS) satellite signal simulator is used to provide realistic test signals for GPS receivers and navigation systems. In this paper, signals arriving the antenna of GPS receiver are analyzed from the viewpoint of simulator design. The estimation methods are focused of which several signal parameters are difficult to determine directly according to existing experiential models due to various error factors. Based on the theory of Artificial Neural Network (ANN), an approach is proposed to simulate signal propagation delay,carrier phase, power, and other parameters using ANN. The architecture of the hardware-in-the-loop test system is given. The ANN training and validation process is described. Experimental results demonstrate that the ANN designed can statistically simulate sample data in high fidelity.Therefore the computation of signal state based on this ANN can meet the design requirement,and can be directly applied to the development of multi-channel GPS satellite signal simulator.展开更多
文摘在基于位置的社交网络中,兴趣点实时推荐数据和用户签到数据存在高稀疏性问题。提出一种基于时间效应的混合推荐模型。通过用户潜在兴趣点数据模型计算用户时间行为影响分数和地理位置影响分数,并用线性统一模型进行处理,选取Top S 个兴趣点作为用户的潜在兴趣点。将用户的潜在签到记录引入基于时间效应的矩阵分解模型中,考虑时间差异性和连续性对推荐结果的影响,在此基础上进行优化求解,提出推荐策略。实验结果表明,与LRT模型、UTE+SE模型相比,该模型的推荐效果较好,其准确率和召回率最高可达0.103 4和0.111 8。
基金Supported by the National Natural Science Foundation of China (No.60027001).
文摘Multi-channel Global Positioning System (GPS) satellite signal simulator is used to provide realistic test signals for GPS receivers and navigation systems. In this paper, signals arriving the antenna of GPS receiver are analyzed from the viewpoint of simulator design. The estimation methods are focused of which several signal parameters are difficult to determine directly according to existing experiential models due to various error factors. Based on the theory of Artificial Neural Network (ANN), an approach is proposed to simulate signal propagation delay,carrier phase, power, and other parameters using ANN. The architecture of the hardware-in-the-loop test system is given. The ANN training and validation process is described. Experimental results demonstrate that the ANN designed can statistically simulate sample data in high fidelity.Therefore the computation of signal state based on this ANN can meet the design requirement,and can be directly applied to the development of multi-channel GPS satellite signal simulator.
