A New Bio-inspired Unsupervised Learning Method
文献类型：会议论文
| 作者 | Kaijian Weng; Guoyuang Liang; Xinyu Wu |
| 出版日期 | 2013 |
| 会议名称 | 6th International Conference on Brain Inspired Cognitive Systems, BICS 2013 |
| 会议地点 | Beijing, China |
| 英文摘要 | Unsupervised learning has been widely used in many areas such as pattern recognition. However, it is usually difficult to acquire accurate representation of pattern within a limited period of time. Unsupervised learning, in general, is likely to be more common in brain than supervised learning. In this paper, we propose a new neural network based unsupervised learning method and evaluate its applications on 1-D and 2-D pattern learning. Our approach is inspired by recent researches on the physiological process of neural connection and brain activity. A bipolar weight scheme based on biological neural connection mechanism is presented. Moreover, we have also noticed the synaptic plasticity of brain plays an important role in learning. A new brain-inspired short-term and long-term scheme is applied in our method to adjust weights during the learning process. Experimental results of learning over 1-D and 2-D patterns demonstrate the proposed method is effective and high-efficiency. |
| 收录类别 | EI |
| 语种 | 英语 |
| 源URL | [http://ir.siat.ac.cn:8080/handle/172644/4612]  |
| 专题 | 深圳先进技术研究院_集成所 |
| 作者单位 | 2013 |
| 推荐引用方式 GB/T 7714 | Kaijian Weng,Guoyuang Liang,Xinyu Wu. A New Bio-inspired Unsupervised Learning Method[C]. 见:6th International Conference on Brain Inspired Cognitive Systems, BICS 2013. Beijing, China. |
入库方式： OAI收割
来源：深圳先进技术研究院
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。