A Large-Scale Hierarchical Structure Knowledge Enhanced Pre-training Framework for Automatic ICD Coding

2021 ◽  
pp. 494-502
Author(s):  
Shi Wang ◽  
Daniel Tang ◽  
Luchen Zhang
2015 ◽  
Vol 11 (A29B) ◽  
pp. 709-710
Author(s):  
Enrique Vázquez-Semadeni ◽  
Gilberto Gómez

AbstractWe discuss the formation of filaments in molecular clouds (MCs) as the result of large-scale collapse in the clouds. We first give arguments suggesting that self-gravity dominates the nonthermal motions, and then briefly describe the resulting structure, similar to that found in molecular-line and dust observations of the filaments in the clouds. The filaments exhibit a hierarchical structure in both density and velocity, suggesting a scale-free nature, similar to that of the cosmic web, resulting from the domination of self-gravity from the MC down to the core scale.


Author(s):  
Masataka Yoshimura ◽  
Kazuhiro Izui

Abstract A large-scale machine system often has a general hierarchical structure. For hierarchical structures, optimization is difficult because many local optima almost always arise, however genetic algorithms that have a hierarchical genotype can be applied to treat such problems directly. Relations between the structural components are analyzed and this information is used to partition the hierarchical structure. Partitioning large-scale problems into sub-problems that can be solved using parallel processed GAs increases the efficiency of the optimization search. The optimization of large-scale systems then becomes possible due to information sharing of Pareto optimum solutions for the sub-problems.


2004 ◽  
Vol 126 (2) ◽  
pp. 217-224 ◽  
Author(s):  
Masataka Yoshimura ◽  
Kazuhiro Izui

A large-scale machine system often has a general hierarchical structure. For hierarchical structures, optimization is difficult because many local optima almost always arise, however genetic algorithms that have a hierarchical genotype can be applied to treat such problems directly. Relations between the structural components are analyzed and this information is used to partition the hierarchical structure. Partitioning large-scale problems into sub-problems that can be solved using parallel processed GAs increases the efficiency of the optimization search. The optimization of large-scale systems then becomes possible due to information sharing of Pareto optimum solutions for the sub-problems.


2017 ◽  
Vol 284 (1862) ◽  
pp. 20171494 ◽  
Author(s):  
Bert Van Bocxlaer

Ecological processes, non-ecological processes or a combination of both may cause reproductive isolation and speciation, but their specific roles and potentially complex interactions in evolutionary radiations remain poorly understood, which defines a central knowledge gap at the interface of microevolution and macroevolution. Here I examine genome scans in combination with phenotypic and environmental data to disentangle how ecological and non-ecological processes contributed to population differentiation and speciation in an ongoing radiation of Lanistes gastropods from the Malawi Basin. I found a remarkable hierarchical structure of differentiation mechanisms in space and time: neutral and mutation-order processes are older and occur mainly between regions, whereas more recent adaptive processes are the main driver of genetic differentiation and reproductive isolation within regions. The strongest differentiation occurs between habitats and between regions, i.e. when ecological and non-ecological processes act synergistically. The structured occurrence of these processes based on the specific geographical setting and ecological opportunities strongly influenced the potential for evolutionary radiation. The results highlight the importance of interactions between various mechanisms of differentiation in evolutionary radiations, and suggest that non-ecological processes are important in adaptive radiations, including those of cichlids. Insight into such interactions is critical to understanding large-scale patterns of organismal diversity.


Sign in / Sign up

Export Citation Format

Share Document