重发统计起薪投票贴～

wwtpcsuper 发表于 2013-4-11 23:27
phd更重要的是职业吧~

啥意思？职业的选择么

看星的馨馨 发表于 2013-4-11 10:29
啥意思？职业的选择么

对呀~~~职业内部感觉大家工资都差不了太多

wwtpcsuper 发表于 2013-4-11 23:39
对呀~~~职业内部感觉大家工资都差不了太多

是指行业，还是职位呢？

看星的馨馨 发表于 2013-4-11 10:42
是指行业，还是职位呢？

行业+职位

wwtpcsuper 发表于 2013-4-11 23:48
行业+职位

行业的话就分金融药厂（IT？）这几大行吧
职位都是analyst类似物 有很大的差别吗？

看星的馨馨 发表于 2013-4-11 10:49
行业的话就分金融药厂（IT？）这几大行吧
职位都是analyst类似物 有很大的差别吗？

data scientist之类的，近几年我们系的很多都去了twitter, GE之类的地方

wwtpcsuper 发表于 2013-4-11 23:51
data scientist之类的，近几年我们系的很多都去了twitter, GE之类的地方

处理哪种类型的数据？

看星的馨馨 发表于 2013-4-11 10:53
处理哪种类型的数据？

具体我也不知道~

大IT公司（FLAGT）看中的往往是处理大数据的经验 比如几个G的data，没法一次性读入到内存里的，要怎么样去应用一些统计方法来处理它们。一般对统计方法要求没那么高，但是编程语言比如SQL，Hadloop，C++之类的要求比较高

highly recommend watching the 1st week's intro video to coursera's "intro to data science".
i think it gave a terrific summary of what data science is, and how it differs from traditional stats/cs.

for stats ppl if you want to enter data science, then one important skill you may not learn in school is data hacking skills.
for cs ppl, maybe a little introductory stats would be helpful.

agree that usually no sophisticated stat modeling is needed , especially in those applied positions. But the google papers do indicate some modeling skills....

it's all about "ETL", "modeling", "visualize/explain". Diff industry/position would have diff requirement on the composition.
the traditional statisticians prob don't do that much ETL, since data is cleaned and prepared by programmers..but lots of modeling
The traditional business analysts prob don't do much modeling, and only averages and sums...but lots of explanations