Search

Citation count is an important indicator for measuring research outputs. There have been numerous studies that have investigated factors affecting citation counts from the perspectives of cited papers and citing papers. In this paper, we focused specifically on citing papers and explored citations sourced from prestigious affiliations in the computer science discipline. The QS World University Rankings was employed to identify prestigious citations, named QS citations. We used the Microsoft Academic Graph, a massive scholarly dataset, and conducted different kinds of analysis between papers with QS citations and those without QS citations. We discovered that papers with QS citations are generally associated with higher total citation counts than those without QS citations. We extended the analysis to authors and journals, and the results indicated that when authors or journals have higher proportions of papers with QS citations, they are usually associated with higher values of the H-index or the Journal Impact Factor respectively. Additionally, papers with QS citations are also associated with a higher Altmetric Attention Score and a higher number of specific types of altmetrics such as tweet counts. [ABSTRACT FROM AUTHOR]

Based on the function approximation theory and the Weierstrass approximation theorem, a novel 3-input neural net activated by a group of products of Bernoulii polynomials (i.e., 3-input Bernoulli polynomial neural net, 3IBPNN) was constructed in this paper. Furthermore, on the basis of the weights-direct-determination (WDD) method and the relationship between the number of hidden-layer neurons and the approximation error of the neural net, three different weights-and-structure-determinattion (WASD) algorithms were built up for the constructed 3IBPNN. Numerical experiment results further prove that all of the 3IBPNNs determined respectively by the three proposed algorithms perform extcellently in training, testing and prediction. [ABSTRACT FROM AUTHOR]