Concept Neural Network — the very beginning

Hi everybody,

This is Haoqi Sun, a man loves neural networks deep into his heart. Here I would like to propose a new kind of neural network — Concept Neural Network (CNN). I hope that with my endless effort, CNN could be very popular and be used in real life applications. Eventually, I would like to start a company selling various of intelligent products equipped with CNNs. The ultimate goal is to make our life more intelligent. (The goal should not be changed once set, unless necessary).

A brief definition of concept (subject to change),

Concept is a learnt (recognized) class of a neural network. Concept is highly connected and hierarchical, where concrete concepts can build abstract concepts.

A brief definition of concept neural network (subject to change),

Concept neural network is one kind of spiking neural networks which use polychronous groups to represent concepts, thus making this network able to understand our world using concepts, as we are.

This is just the beginning of this fantastic story. I’ll try to explain concept neural networks gradually. Other things including network structure, learning, toolboxes and papers will follow.

This is my contact info,
Email: [e-mail]
Addr: Tongji University, Siping Rd No. 1239, Shanghai, China
You can find my CV in http://www.tongji.edu.cn/~qiliu/people.html

Welcome to drop a comment or email if you are also interested in this strange idea!

Best Regards,
Haoqi Sun

PS. CSDN上的文章会和我在blog.com上的文章同步更新。

blog.com上的博客地址：http://conceptnn.blog.com/