Earticle

초록

영어

Nowadays, the big success of deep learning makes artificial neural network becoming a hot topic once again, and the size of neural networks’ structure is a key visual cue for structured learning. The greater network may get the study task done well, while it may increase network computation overhead easier and cost more. Hence, network construction is an important issue, as well as a difficult problem. In this paper, we proposed a novel sensitivity-based adaptive architecture pruning algorithm for Madalines. The algorithm establishes a pruning measure based on the network sensitivity to its structure variation and a minimal disturbance principle. The measure can be used to evaluate the performance loss due to its structure changes more or less. And the loss can be compensated by relearning. Thus, the new adaptive pruning mechanism is developed with measuring, pruning, and compensating. The simulation experimental results based on some benchmark data demonstrate that the pruning measure is rationality and the new algorithm is effective.

목차

Abstract
 1. Introduction
 2. Preliminaries
  2.1. Madalines Model
  2.2. Madalines Sensitivity
 3. Sensitivity of Madalines Based on Architecture
  3.1. Sensitivity Definition Based on Architecture
  3.2. Sensitivity Computation based on Architecture
 4. Pruning Measure Based on Sensitivity
 5. The Pruning Algorithm
 6. Experimental Verifications
 7. Conclusion
 Acknowledgements
 References

키워드

저자

• Sai Ji [ Jiangsu Engineering Center of Network Monitoring, Nanjing University of Information Science and Technology, Nanjing 210044，China ]
• Ping Yang [ Jiangsu Engineering Center of Network Monitoring, Nanjing University of Information Science and Technology, Nanjing 210044，China ]
• Shuiming Zhong [ Jiangsu Engineering Center of Network Monitoring, Nanjing University of Information Science and Technology, Nanjing 210044，China ]
• Jin Wang [ School of Information Engineering, Yangzhou University, Yangzhou 225009, China ]
• Jeong-Uk Kim [ Department of Energy Grid, Sangmyung University, Seoul, Korea ]

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  International Journal of Hybrid Information Technology

• 간기

  격월간

• pISSN

  1738-9968

• 수록기간

  2008~2016

• 십진분류

  KDC 505 DDC 605