相关实验视频
Updated: Jun 6, 2025

Revealing Neural Circuit Topography in Multi-Color
Published on: November 14, 2011
为可扩展的图形神经网络提供Coarsening框架
Shengzhong Zhang1, Yimin Zhang2, Bisheng Li3
1Fudan University, 220 Handan Road, Shanghai, 200433, China.
图形批量凝聚 (GBC) 提供了一种在大型数据集上训练图形神经网络 (GNN) 的新方法. 这种方法避免了随机抽样,提高了准确性,减少了训练时间和内存使用.
科学领域:
- 图形神经网络 (GNN) 是一个神经网络.
- 在图表上进行机器学习.
- 可扩展的图形分析.
背景情况:
- 将图形神经网络 (GNN) 扩展到大型图形是具有挑战性的,因为邻里爆炸现象.
- 现有的以采样为基础的小批量方法 (以节点为基础,以层为基础,以子图为基础的采样) 带来了开销和不一致的性能.
- 在GNN培训中随机抽样可能是低效的,并影响模型的有效性.
研究的目的:
- 引入GBC (图表批量缩),这是一个可扩展的GNN培训的新框架.
- 为提供一个通用的解决方案,以促进在大型图形上训练任意的GNN模型.
- 为了克服在GNN培训中随机抽样的局限性.
主要方法:
- 图表批量粗化 (GBC) 预先将输入图表处理成更小的子图,用于小型批量训练.
- 该框架采用了利用标签传播的图形分解方法.
- 使用了一种专门为GNN培训而设计的新型图形粗化算法.
主要成果:
- GBC完全避免随机抽样,简化了培训过程.
- 该框架不需要对现有的GNN模型或其超参数进行修改.
- 经验结果显示,在各种图表尺度中,精度,训练时间缩短和内存使用率较低的性能优越.
结论:
- 图表批量硬化 (GBC) 为可扩展的GNN培训提供了一种有效和可泛化的方法.
- 与传统采样技术相比,该方法显著提高了效率和性能.
- 对于将GNN应用于大规模图形数据,GBC提供了一个有前途的方向.
更多相关视频
12:27Large-scale Reconstructions and Independent, Unbiased Clustering Based on Morphological Metrics to Classify Neurons in Selective Populations
Published on: February 15, 2017
03:31Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
相关概念视频
Vector Algebra: Graphical Method
We use the laws of geometry to construct resultant vectors, followed by trigonometry to find vector magnitudes and directions. For a geometric construction of the sum of two vectors in a plane, we follow the parallelogram rule. Suppose two vectors are at arbitrary positions. Translate either one of...
Parallel Processing
End Point Prediction: Gran Plot
For potentiometric titration, the Gran plot is created by plotting...
Design Example: Aggregate Gradation
The grading, or particle-size distribution, of sand is determined using sieve analysis, with standard sizes ranging from 150 μm to 10 mm (ASTM No. 100 sieve to 3⁄8 in. sieve). Sand is...
Multiple Bar Graph
Each bar or column in the multiple bar graph represents a data value. These graphs are used primarily in interrelating two or more sets of data. The categories of different kinds of data are listed along the horizontal or x-axis, whereas...
Reducing Line Loss
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...