150.txt

来自「This complete matlab for neural network」· 文本 代码 · 共 32 行

发信人: helloboy (hello), 信区: DataMining
标  题: Re: Problem
发信站: 南京大学小百合站 (Sat Dec 29 10:41:17 2001), 站内信件

But it's a non-trival process.
I need much time to read all the books.But at the same time I need 
to write articles for my doctoral degree articles. I can't see all the 
materials. But I'm afraid my articles will become invalid because my
idea is the same with someone else.

【 在 daniel (飞翔鸟) 的大作中提到: 】
: 【 在 roamingo (漫步鸥) 的大作中提到: 】
: : I suggest that one should be familiar with the literature before
: : he/she starts to made some new advances.  However, for your question,
: : the best way I believe is to ask a top expert, e.g. Jiawei Han, to 
: : validate whether your idea is new.  A good DIY way I think is
: : to trace the citation threads with "http://www.researchindex.org/"
: : However, you should still know some of the top papers in your field
: : to start with.
: I will not encourage you to query experts such as Han Jiawei because there 
: are always such kinds of questions badering them. A good solution for you 
: is to read prestigious journals or conference proceedings involving your 
: area at first. For data mining, the journal DMKD, ML, DKE, and IEEE TKDE, 
: and the proceedings of KDD, ICDM, SIGMOD, and VLDB, should not be missed. 
: Also you should not miss SIGKDD Explorations. You d better work on
: your idea after mastering the state-of-the-art of the area. Otherwise your
: time may have great chances to be wasted.


--
※ 来源:．南京大学小百合站 bbs.nju.edu.cn．[FROM: 202.38.215.15]