Several algorithms have been proposed to learn to rank entities modeled as feature vectors, based on relevance feedback. However, these algorithms do not model network connections or relations between entities. Meanwhile, Pagerank and variants find the stationary distribution of a reasonable but arbitrary Markov walk over a network, but do not learn from relevance feedback. We present a framework for ranking networked entities based on Markov walks with parameterized conductance values associated with the network edges. We propose two flavors of conductance learning problems in our framework. In the first setting, relevance feedback comparing node-pairs hints that the user has one or more hidden preferred communities with large edge conductance, and the algorithm must discover these communities. We present a constrained maximum entropy network flow formulation whose dual can be solved efficiently using a cutting-plane approach and a quasi-Newton optimizer. In the second setting, edges have types, and relevance feedback hints that each edge type has a potentially different conductance, but this is fixed across the whole network. Our algorithm learns the conductances using an approximate Newton method.
Energy levels, lifetimes and wavefunction compositions have been computed for all levels of odd parity 4s24p5 ground configuration as well as 4s4p6 and 4s24p44d even parity excited configurations in Br-like Sr IV, Y V, Zr VI, Nb VII and Mo VIII. Transition probabilities, oscillator strengths and line strengths for the electric dipole (E1) transition from the 4s24p5 configuration have been obtained using the multiconfiguration Dirac–Fock approach. Correlations within the n = 4 complex, Breit and quantum electrodynamics effects have been included. We make a detailed comparison of our results with those of other numerical methods and experiments to assess the quality of our results. Good agreement is observed between our results and those obtained using different approaches confirm the quality of our results. Further, we have also predicted new atomic data that were not available so far and are yet to be observed.
Energy levels, lifetimes, and wavefunction compositions have been calculated for all levels of odd parity 3s23p5 ground configuration as well as 3s3p6 and 3s23p43d even parity excited configurations in highly charged Cl-like tungsten ion. Transition probabilities, oscillator strengths, and line strengths for E1, E2, M1, and M2 transitions have been obtained using the fully relativistic multiconfiguration Dirac–Fock (MCDF) approach including the correlations within the n = 3 complex, some n = 3 → n = 4 single and double excitations and Breit and quantum electrodynamics effects. For comparison from our calculated energy levels, we have also calculated the energy levels by using the fully relativistic flexible atomic code (FAC). The validity of the method is assessed by comparison with previously published experimental and theoretical data. The excellent agreement observed between our calculated results and those obtained using different approaches confirm the accuracy of our results. Additionally, we have predicted some new atomic data for W57+ that are not available so far and may be important for plasma diagnostic analysis in fusion plasma.
We report calculations of energy levels and oscillator strengths for transitions in W XL, undertaken with the general-purpose relativistic atomic structure package (grasp) and flexible atomic code (fac). Comparisons are made with existing results and the accuracy of the data is assessed. Discrepancies with the most recent results of S. Aggarwal et al. [Can. J. Phys. 91 (2013) 394] are up to 0.4 Ryd and up to two orders of magnitude for energy levels and oscillator strengths, respectively. Discrepancies for lifetimes are even larger, up to four orders of magnitude for some levels. Our energy levels are estimated to be accurate to better than 0.5% (i.e. 0.2 Ryd), whereas results for oscillator strengths and lifetimes should be accurate to better than 20%. Energy levelsFor our calculations we have adopted the same grasp code as employed by AJM [4]. Similarly, we have also used the option of extended average level (EAL), in which a weighted (proportional to 2j+1) trace of the Hamiltonian matrix is minimised. This produces a compromise set of orbitals describing closely-lying states with moderate accuracy, and generally yields results comparable to other options, such as average level (AL), as noted by Aggarwal et al. for several ions of Kr [7] and Xe [8]. Furthermore, to assess the accuracy of our results, we have also employed the Flexible Atomic Code (fac) of Gu [9], available from the website http://sprg.ssl.berkeley.edu/∼mfgu/fac/. This is also a fully relativistic code which provides a variety of atomic parameters, and yields results for energy levels and oscillator strengths comparable to grasp, as
Motived by the recent measurement of transition lines for Ne-like Hf and W, we have reported atomic data in the form of multiconfiguration Dirac–Fock transition energies and wavefunction compositions of 209 levels belonging to the configurations 2s22p6, 2s22p5ns (n = 3, 4, 5, 6, 7), 2s22p5np (n = 3, 4, 5, 6, 7), 2s22p5nd (n = 3, 4, 5, 6, 7), 2s22p5nf (n = 4, 5), 2s22p55g, 2s2p6ns (n = 3, 4, 5), 2s2p6np (n = 3, 4, 5), 2s2p6nd (n = 3, 4, 5), 2s2p6nf (n = 4, 5), and 2s2p65g of Hf LXIII, Ta LXIV, W LXV, and Re LXVI. Radiative rates, oscillator strengths, transition wavelengths, and line strengths have been calculated for ground state electric dipole (E1) transition among these levels. These values were obtained using GRASP (general-purpose relativistic atomic structure package) code, which includes Breit and QED effects along with Dirac–Fock potential and second-order Coulomb interaction. We have compared our results with the data compiled using FAC (flexible atomic code) and also with the recent results available in the literature. The accuracy of the data is assessed. We predict new energy levels, oscillator strength, and transition probability data, where no other theoretical or experimental results are available, which will form the basis for future experimental work.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.