In order to supply better accordance for mod eling and simulation of complex networks, a new degree dependence entropy (DDE) descriptor is proposed to describe the degree dependence relationship and corre sponding c...In order to supply better accordance for mod eling and simulation of complex networks, a new degree dependence entropy (DDE) descriptor is proposed to describe the degree dependence relationship and corre sponding characteristic in this paper. First of all, degrees of vertices and the shortest path lengths between all pairs of ,ertices are computed. Then the degree dependence matrices under different shortest path lengths are con structed. At last the DDEs are extracted from the degree dependence matrices. Simulation results show that the DDE descriptor can reflect the complexity of degree dependence relationship in complex networks; high DDE indicates complex degree dependence relationship; low DDE indicates the opposite one. The DDE can be seen as a quantitative statistical characteristic, which is meaningful for networked modeling and simulation.展开更多
针对室内环境热舒适度评价,为解决影响PMV(predicted mean vote)指标的各因素之间复杂的非线性关系,利用核主成分分析KPCA(kernel principal component analysis)的非线性映射方法,对输入变量进行特征提取,以消除各因素之间的非线性关系...针对室内环境热舒适度评价,为解决影响PMV(predicted mean vote)指标的各因素之间复杂的非线性关系,利用核主成分分析KPCA(kernel principal component analysis)的非线性映射方法,对输入变量进行特征提取,以消除各因素之间的非线性关系,然后利用遗传神经网络GNN(genetic neural network)进行融合评价。对比GNN和KPCA+GNN的仿真评价结果可知:对于该室内热环境舒适度融合评价问题,KPCA能提取影响PMV指标的主要因素成分,KPCA+GNN是有效的预测方法。展开更多
基金supported by the National Natural Science Foundation of China(Grants Nos.61174156,61273189,71073172,61174035,61203140)
文摘In order to supply better accordance for mod eling and simulation of complex networks, a new degree dependence entropy (DDE) descriptor is proposed to describe the degree dependence relationship and corre sponding characteristic in this paper. First of all, degrees of vertices and the shortest path lengths between all pairs of ,ertices are computed. Then the degree dependence matrices under different shortest path lengths are con structed. At last the DDEs are extracted from the degree dependence matrices. Simulation results show that the DDE descriptor can reflect the complexity of degree dependence relationship in complex networks; high DDE indicates complex degree dependence relationship; low DDE indicates the opposite one. The DDE can be seen as a quantitative statistical characteristic, which is meaningful for networked modeling and simulation.
文摘针对室内环境热舒适度评价,为解决影响PMV(predicted mean vote)指标的各因素之间复杂的非线性关系,利用核主成分分析KPCA(kernel principal component analysis)的非线性映射方法,对输入变量进行特征提取,以消除各因素之间的非线性关系,然后利用遗传神经网络GNN(genetic neural network)进行融合评价。对比GNN和KPCA+GNN的仿真评价结果可知:对于该室内热环境舒适度融合评价问题,KPCA能提取影响PMV指标的主要因素成分,KPCA+GNN是有效的预测方法。