China 
Africa 
Explore chapters and articles related to this topic
Analysis: Four Levels for Validation
Published in Tamara Munzner, Visualization Analysis and Design, 2014
cause the data type in fact is a more general graph with specialized constraints on its structure. They discuss conditions for which the data type is a true tree, a multitree, or a directed acyclic graph. They map the domain problem of recognizing nuclear family structure into an abstract task of determining subgraph structure. At the third level of the model, they discuss the strengths and weaknesses of several visual encoding idiom design choices, including using connection, containment, adjacency and alignment, and indentation. They present in passing two more specialized encoding idioms, fractal nodelink and containment for free trees, before presenting in detail their main proposal for visual encoding. They also carefully address interaction idiom design, which also falls into the third level of the model. At the fourth level of algorithm design, they concisely cover the algorithmic details of dual-tree layout. Three validation methods are used in this paper, shown in Figure 4.7. There is the immediate justification of encoding and interaction idiom design decisions in terms of established principles, and the downstream method of a qualitative discussion of result images and videos. At the abstraction level, there is the downstream informal testing of a system prototype with a target user to collect anecdotal evidence.
Research on the Preliminary Prediction of Nuclear Core Design Based on Machine Learning
Published in Nuclear Technology, 2022
Jichong Lei, Zhenping Chen, Jiandong Zhou, Chao Yang, Changan Ren, Wei Li, Chao Xie, Zining Ni, Gan Huang, Leiming Li, Jinsen Xie, Tao Yu
The decision tree algorithm is easy to understand without much professional background. It can process the data of multiple attributes at the same time and predict with high precision in a short time. The decision tree model is a hierarchical model of a binary tree or multitree structure based on the principle of minimizing loss function, which could realize complex hierarchical decisions or multilevel decisions. The model is shown in Fig. 1. C4.5 is an algorithm of decision trees.13 The loss function of decision tree learning is mainly based on the maximum likelihood estimation of regularization, and the specific regularization loss function is shown as follows: