j******e
发帖数: 232 | |
s*******y
发帖数: 558 | 2 解释一下
【在 j******e 的大作中提到】
: 阻碍了计算机应有的快速发展
|
t*******l
发帖数: 3662 | 3
CS本来也不是指望着在学校发展的，CS的前沿发展都在公司
【在 j******e 的大作中提到】
: 阻碍了计算机应有的快速发展
|
S*F
发帖数: 59 | 4 公司那帮人还不都是学校出去的？ :-)
指挥棒不一样，做得就不一样，公司要利润，学校要funding....
【在 t*******l 的大作中提到】
:
: CS本来也不是指望着在学校发展的，CS的前沿发展都在公司
|
wy
发帖数: 14511 | 5 ft.
【在 t*******l 的大作中提到】
:
: CS本来也不是指望着在学校发展的，CS的前沿发展都在公司
|
s*******y
发帖数: 558 | 6 你要说具体的application 当然是公司开发的fancy东西多
这是相对短期利润驱动的
可是学校做的更多的是基础研究 我们每天更多做的是数学公式推导
定理证明 算法研究 短期来看产生不了什么利润 甚至做到头发现
是negative result 但是对于长期的发展奠定些基础
感觉你对cs的认识太肤浅叻
【在 t*******l 的大作中提到】
:
: CS本来也不是指望着在学校发展的，CS的前沿发展都在公司
|
t*******l
发帖数: 3662 | 7
you know little about industry's research.
check out intel, hp, nec and ibm's patents, you will know
they not only do 'fancy applications', but also do research.
【在 s*******y 的大作中提到】
: 你要说具体的application 当然是公司开发的fancy东西多
: 这是相对短期利润驱动的
: 可是学校做的更多的是基础研究 我们每天更多做的是数学公式推导
: 定理证明 算法研究 短期来看产生不了什么利润 甚至做到头发现
: 是negative result 但是对于长期的发展奠定些基础
: 感觉你对cs的认识太肤浅叻
|
D*****r
发帖数: 183 | 8 他们做的research都是能够马上/短期内能拿钱那种，
真正的research，主要基地还是在学术界，这也是industry/academic的区别所在。
但是现在很多公司都倾向于把钱用来与学校合作而不是自己设立研究中心。
反正两种方式各有利弊。
【在 t*******l 的大作中提到】
:
: you know little about industry's research.
: check out intel, hp, nec and ibm's patents, you will know
: they not only do 'fancy applications', but also do research.
|
T********r
发帖数: 6210 | 9 这个不能苟同, 我知道IBM有很多research都是超前的, 没什么短期效益可言.
【在 D*****r 的大作中提到】
: 他们做的research都是能够马上/短期内能拿钱那种，
: 真正的research，主要基地还是在学术界，这也是industry/academic的区别所在。
: 但是现在很多公司都倾向于把钱用来与学校合作而不是自己设立研究中心。
: 反正两种方式各有利弊。
|
x***u
发帖数: 336 | 10 not any more....
【在 T********r 的大作中提到】
: 这个不能苟同, 我知道IBM有很多research都是超前的, 没什么短期效益可言.
|
|
|
D*****r
发帖数: 183 | 11 这得看你怎么看，学科交叉还是有很多潜力可以挖掘的。
按你这种说法，很多学科基础研究都不必作了，因为真正的基础学科如数理化生
方面如果没有突破性进展，气它学科也很难说有什么新东西。 |
D*****r
发帖数: 183 | 12 CS+statistics，CS+Bioinformatics
.. |
D*****r
发帖数: 183 | 13 交叉学科基本上都是酱紫，
前者无论如何都不能说是
纯applicatino-oriented. |
S******o
发帖数: 111 | 14 Yeah, Statistics has been so extensively used in so many fields, for example,
networks, information retrieval, information extraction, machine learning,
etc. To me I think it's kinda like a tool for computer science rather than a
research object.
【在 D*****r 的大作中提到】
: 交叉学科基本上都是酱紫，
: 前者无论如何都不能说是
: 纯applicatino-oriented.
|
x***u
发帖数: 336 | 15 I don't think so. Many areas like you said, machine learning, information
retrieval, are indeed research areas in computer science. If not, what's your
definition for computer science, and who can really define computer science
anyway?!
And I don't agree with earlier people. There are tons of open problems in
foundamental computer science, such as complexity theory, algorithms, etc. I
hope people be open, and learn to appreciate others' work.
example,
【在 S******o 的大作中提到】
: Yeah, Statistics has been so extensively used in so many fields, for example,
: networks, information retrieval, information extraction, machine learning,
: etc. To me I think it's kinda like a tool for computer science rather than a
: research object.
|
s***x
发帖数: 34 | 16 Not really, there are lots of theoretical research going on at places
like IBM Research and MS Research. Many thoery papers are published by
these people.
【在 D*****r 的大作中提到】
: 他们做的research都是能够马上/短期内能拿钱那种，
: 真正的research，主要基地还是在学术界，这也是industry/academic的区别所在。
: 但是现在很多公司都倾向于把钱用来与学校合作而不是自己设立研究中心。
: 反正两种方式各有利弊。
|
x***u
发帖数: 336 | 17 Open problems mean currently there are no solutions. But in the future people
may solve that, positively or negatively. The research papers in computer
theory and algorithms published each year solve such problems. Well although
not many are as significant as P=NP. But I would many of them are meaningful
problems.
your
science
I
learning,
than a |
x***u
发帖数: 336 | 18 oh, me too. :)
maybe my words earlier were a little strong. But indeed I think there are
misunderstandings between systems people and theory people. Both sides kind of
look down the other side. :( Some theory people work on artificial problems
and publish papers that stay in library forever. On the other hand, system
people just use simple heuristics without any rigorous analysis. This is not
surprising though, because firstly the theory are not powerful enough to
analyze everything, and nobody |
j*********t
发帖数: 11 | 19 本质上是因为：
1) 很多实际问题，没有什么RESEARCH可做，工程含量更高。
2) 管用的RESEARCH太难做了。
只顾一头，就容易的多。所以很多人只做PAPER，根本也不想用到实际中去；
而另外一些人只做项目，WORKABLE 才是硬道理。
两边都想做的人也有，做好的，声名远播。做不出来的，谋生都困难。
of
I
【在 x***u 的大作中提到】
: oh, me too. :)
: maybe my words earlier were a little strong. But indeed I think there are
: misunderstandings between systems people and theory people. Both sides kind of
: look down the other side. :( Some theory people work on artificial problems
: and publish papers that stay in library forever. On the other hand, system
: people just use simple heuristics without any rigorous analysis. This is not
: surprising though, because firstly the theory are not powerful enough to
: analyze everything, and nobody
|
f*****r
发帖数: 229 | 20 Yes, system guys always use heuristics. But I found that if you have very
clear understanding of theory, you will have more elegant and general
heuristic, and not only some tricks for your workload and dataset. Such as, if
you have deep understanding of queue theory, you can design better scheme,
even though for desiging a monitor to collect critical measured data. Also,
if you have good understanding to stochastic process and prediction, you can
have deep understanding how prediction works in
【在 x***u 的大作中提到】
: oh, me too. :)
: maybe my words earlier were a little strong. But indeed I think there are
: misunderstandings between systems people and theory people. Both sides kind of
: look down the other side. :( Some theory people work on artificial problems
: and publish papers that stay in library forever. On the other hand, system
: people just use simple heuristics without any rigorous analysis. This is not
: surprising though, because firstly the theory are not powerful enough to
: analyze everything, and nobody
|
