48.txt

This complete matlab for neural network

发信人: GzLi (笑梨), 信区: DataMining
标  题: [合集]ML书里说ANN的研究分两个团体
发信站: 南京大学小百合站 (Sat Jan 11 21:56:30 2003)

acat (考完了还要干活:() 于Tue Jan  7 22:35:36 2003)
提到：

一个使用之来模拟生物学习过程
一个则脱离生物的过程

第一种的研究是不是太薄弱了。


daniel (飞翔鸟) 于Wed Jan  8 01:02:29 2003)
提到：

  NN is a very complex community. Mitchell just meant two main motivations

of NN researchers. If you peer at the community you may find researchers

with different background are working with different "NN", e.g. these

from mathematics are interested in stability or convergence, these from 

computer science focus on algorithms and generalisation, these

from automation work on neural control, these from electronics investigate

neural circuit, these from physiology want to get illumination from 

artificial NN on biological NN, these from psychology wish explain their

experimental behavior data with NN models, ...
 

【 在 acat (考完了还要干活:() 的大作中提到: 】
: 一个使用之来模拟生物学习过程
: 一个则脱离生物的过程
: 
: 第一种的研究是不是太薄弱了。


strawman (独上江楼思渺然) 于Wed Jan  8 09:49:21 2003)
提到：

【 在 daniel (飞翔鸟) 的大作中提到: 】
:   NN is a very complex community. Mitchell just meant two main motivations
: of NN researchers. If you peer at the community you may find researchers
: with different background are working with different "NN", e.g. these
: from mathematics are interested in stability or convergence, these from 
: computer science focus on algorithms and generalisation, these
                            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
那咱们的着眼点是不是太浅了些？要设计一个好的算法使其有好的泛化能力，毕竟是要
以数学为基础的啊，要关心NN的stability, convergence,还有capacity.

再请教：我记得有一篇阐述BP NN 是 universal approximator的文章，不知道在网上
能否找到？能否提示一二？
  
: from automation work on neural control, these from electronics investigate
: neural circuit, these from physiology want to get illumination from 
: artificial NN on biological NN, these from psychology wish explain their
: experimental behavior data with NN models, ...
: 【 在 acat (考完了还要干活:() 的大作中提到: 】


daniel (飞翔鸟) 于Wed Jan  8 12:34:48 2003)
提到：

【 在 strawman (独上江楼思渺然) 的大作中提到: 】
: 【 在 daniel (飞翔鸟) 的大作中提到: 】
:                             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
: 那咱们的着眼点是不是太浅了些？要设计一个好的算法使其有好的泛化能力，毕竟是要
: 以数学为基础的啊，要关心NN的stability, convergence,还有capacity.

I don;t know why you think it is "shallow". They are just different aspects
of an object. Do you think designing algorithms does not require mathematics?
Be aware that nothing can live without mathematics. 

I don't know how do you care the stability and convergence, do you have tried 
to prove the stability or convergency of any practical NN model? For most 
computer scientists, obtaining a well-working algorithm is good enough.
In fact, most properties of most practical NN models have not been proved
at present.

: 再请教：我记得有一篇阐述BP NN 是 universal approximator的文章，不知道在网上
: 能否找到？能否提示一二？

not one. many papers on this topic. one way is from Kolmogrov's work, which 
initials debating for years. The other way is more constructive. But both
are quite difficult to understand for pure computer students. In fact, for
most NN researchers, be aware of the conclusion is enough.
I don't think any of those papers can be found on the internet because they
were worked out at the begining of 1990s.



bigmeat (笑笑生) 于Wed Jan  8 12:48:06 2003)
提到：

【 在 daniel 的大作中提到: 】
: In fact, most properties of most practical NN models have not been proved
: at present.
能不能说具体一点？


strawman (独上江楼思渺然) 于Wed Jan  8 22:12:03 2003)
提到：

【 在 daniel (飞翔鸟) 的大作中提到: 】
: 【 在 strawman (独上江楼思渺然) 的大作中提到: 】
: I don;t know why you think it is "shallow". They are just different aspects
: of an object. Do you think designing algorithms does not require mathematics?
: Be aware that nothing can live without mathematics. 
: I don't know how do you care the stability and convergence, do you have tried 
: to prove the stability or convergency of any practical NN model? For most 
: computer scientists, obtaining a well-working algorithm is good enough.
: In fact, most properties of most practical NN models have not been proved
: at present.
     呵呵，我没有去证明这类的问题。我只是想，利用数学家的这些成果来设计算法。
那既然NN的许多性质并没有被证明，那么这些算法well-working的本质又是什么呢？
这也许又是数学家的事了。
: not one. many papers on this topic. one way is from Kolmogrov's work, which 
: initials debating for years. The other way is more constructive. But both
: are quite difficult to understand for pure computer students. In fact, for
: most NN researchers, be aware of the conclusion is enough.
: I don't think any of those papers can be found on the internet because they
: were worked out at the begining of 1990s.
: (以下引言省略 ... ...)


yinxucheng (yxc) 于Thu Jan  9 13:35:33 2003)
提到：

我觉得进行应用科学研究与学习，主要有三种方式：

（1）偏理论
主要是利用数学的东西，如定义、定理和证明等；在人工神经网络的里面，主要指网络的
数学机理、收敛等。

（2）偏抽象
主要是根据已有数学定理（或没有），对存在的现象提出相应的模型，然后进行分析，
并设计实验验证、改进；在人工神经网络的里面，主要指网络模型与结构、学习方法等。
好像，2001年Science上的文章“Machine Learning of Science: State of the Art and
 Future Prospects”就是以这种“偏抽象”的思路来论述的。

（3）偏应用
主要是根据已有的模型，给出需求与分析，设计实现，并应用到具体的实践中；在人工神
经网络的里面，主要指针对具体的网络结构与学习方法，进行算法设计与实现，当然，在
应用的过程中，需要对相应的方法进行改良。像应用BP网络进行模式识别就是此类方式。

我认为，对于计算机学科（包括其它工科）的学生来说，首先应该是进行“偏应用”的研
究与学习；之后，如果对于某一个具体的点（小方向）很感兴趣，可以进行一定的“偏抽
象”学习；至于“偏理论”方式，除非你的数学功底非常好，否则最好不要去尝试。

不知上面的思路有什么问题，请各位大虾指点指点！

【 在 acat 的大作中提到: 】
: 一个使用之来模拟生物学习过程
: 一个则脱离生物的过程
: 
: 第一种的研究是不是太薄弱了。