suggestion on geospatial data? (转载)# DataSciences - 数据科学
P*e
1 楼
等三天以后
还在 就可以拿包子了。
一是 我和bro对话
二是 我和TTTTBB对话
算两例：）
另外，也请这里的id们到中学时代玩玩，middle schoo版 很好玩的。
还在 就可以拿包子了。
一是 我和bro对话
二是 我和TTTTBB对话
算两例：）
另外，也请这里的id们到中学时代玩玩，middle schoo版 很好玩的。
c*z
2 楼
【 以下文字转载自 Statistics 讨论区 】
发信人: chaoz (面朝大海，吃碗凉皮), 信区: Statistics
标 题: suggestion on geospatial data?
发信站: BBS 未名空间站 (Mon Jun 30 12:48:25 2014, 美东)
Hi all,
I am looking at a project related to geospatial data. The sample sizes are
small and samples are highly correlated.
Can anyone give some suggestion on how to deal with these kind of data?
Thanks a lot!
发信人: chaoz (面朝大海，吃碗凉皮), 信区: Statistics
标 题: suggestion on geospatial data?
发信站: BBS 未名空间站 (Mon Jun 30 12:48:25 2014, 美东)
Hi all,
I am looking at a project related to geospatial data. The sample sizes are
small and samples are highly correlated.
Can anyone give some suggestion on how to deal with these kind of data?
Thanks a lot!
F*y
3 楼
好,加油!:)
【在 P**********e 的大作中提到】
: 等三天以后
: 还在 就可以拿包子了。
: 一是 我和bro对话
: 二是 我和TTTTBB对话
: 算两例：）
: 另外，也请这里的id们到中学时代玩玩，middle schoo版 很好玩的。
【在 P**********e 的大作中提到】
: 等三天以后
: 还在 就可以拿包子了。
: 一是 我和bro对话
: 二是 我和TTTTBB对话
: 算两例：）
: 另外，也请这里的id们到中学时代玩玩，middle schoo版 很好玩的。
d*n
4 楼
$100/hr
b*o
5 楼
呵呵,还有bonus的
【在 P**********e 的大作中提到】
: 等三天以后
: 还在 就可以拿包子了。
: 一是 我和bro对话
: 二是 我和TTTTBB对话
: 算两例：）
: 另外，也请这里的id们到中学时代玩玩，middle schoo版 很好玩的。
【在 P**********e 的大作中提到】
: 等三天以后
: 还在 就可以拿包子了。
: 一是 我和bro对话
: 二是 我和TTTTBB对话
: 算两例：）
: 另外，也请这里的id们到中学时代玩玩，middle schoo版 很好玩的。
T*u
6 楼
你把背景说说
P*e
7 楼
谢谢：）
【在 F*****y 的大作中提到】
: 好,加油!:)
【在 F*****y 的大作中提到】
: 好,加油!:)
z*e
8 楼
沙发太贵了
我的开价是：
aud100/h
我的开价是：
aud100/h
P*e
9 楼
nod nod
【在 b*o 的大作中提到】
: 呵呵,还有bonus的
【在 b*o 的大作中提到】
: 呵呵,还有bonus的
c*z
10 楼
Some background:
This project is related to oil and gas drilling, e.g. where to drill and how
to drill.
I am not a domain expert on it and this is about all the information I have.
Also, the data is geospatial, there are few data points and they are highly
correlated.
This project is related to oil and gas drilling, e.g. where to drill and how
to drill.
I am not a domain expert on it and this is about all the information I have.
Also, the data is geospatial, there are few data points and they are highly
correlated.
c*z
11 楼
Is 100$/hr the rate you guys charge? Actually not bad. :)
c*t
12 楼
这不是data问题吧。不如去土木工程版上吼一声
how
have.
highly
【在 c***z 的大作中提到】
: Some background:
: This project is related to oil and gas drilling, e.g. where to drill and how
: to drill.
: I am not a domain expert on it and this is about all the information I have.
: Also, the data is geospatial, there are few data points and they are highly
: correlated.
how
have.
highly
【在 c***z 的大作中提到】
: Some background:
: This project is related to oil and gas drilling, e.g. where to drill and how
: to drill.
: I am not a domain expert on it and this is about all the information I have.
: Also, the data is geospatial, there are few data points and they are highly
: correlated.
l*m
13 楼
first, formulate a supervised learning.
then, discuss different models and algorithms.
how
have.
highly
【在 c***z 的大作中提到】
: Some background:
: This project is related to oil and gas drilling, e.g. where to drill and how
: to drill.
: I am not a domain expert on it and this is about all the information I have.
: Also, the data is geospatial, there are few data points and they are highly
: correlated.
then, discuss different models and algorithms.
how
have.
highly
【在 c***z 的大作中提到】
: Some background:
: This project is related to oil and gas drilling, e.g. where to drill and how
: to drill.
: I am not a domain expert on it and this is about all the information I have.
: Also, the data is geospatial, there are few data points and they are highly
: correlated.
e*y
14 楼
kriging or multiple points kriging
b*e
15 楼
kriging seems familiar. Google and refresh the knowledge.
Interpolates a raster surface from points using kriging.
Surface interpolation tools create a continuous (or prediction) surface from
sampled point values.
Visiting every location in a study area to measure the height, concentration
, or magnitude of a phenomenon is usually difficult or expensive. Instead,
you can measure the phenomenon at strategically dispersed sample locations,
and predicted values can be assigned to all other locations. Input points
can be either randomly or regularly spaced or based on a sampling scheme.
The continuous surface representation of a raster dataset represents some
measure, such as the height, concentration, or magnitude (for example,
elevation, acidity, or noise level). Surface interpolation tools make
predictions from sample measurements for all locations in an output raster
dataset, whether or not a measurement has been taken at the location.
There are a variety of ways to derive a prediction for each location; each
method is referred to as a model. With each model, there are different
assumptions made of the data, and certain models are more applicable for
specific data—for example, one model may account for local variation better
than another. Each model produces predictions using different calculations.
The interpolation tools are generally divided into deterministic and
geostatistical methods.
The geostatistical methods are based on statistical models that include
autocorrelation (the statistical relationship among the measured points).
Because of this, geostatistical techniques not only have the capability of
producing a prediction surface but also provide some measure of the
certainty or accuracy of the predictions.
Kriging is a geostatistical method of interpolation.
Interpolates a raster surface from points using kriging.
Surface interpolation tools create a continuous (or prediction) surface from
sampled point values.
Visiting every location in a study area to measure the height, concentration
, or magnitude of a phenomenon is usually difficult or expensive. Instead,
you can measure the phenomenon at strategically dispersed sample locations,
and predicted values can be assigned to all other locations. Input points
can be either randomly or regularly spaced or based on a sampling scheme.
The continuous surface representation of a raster dataset represents some
measure, such as the height, concentration, or magnitude (for example,
elevation, acidity, or noise level). Surface interpolation tools make
predictions from sample measurements for all locations in an output raster
dataset, whether or not a measurement has been taken at the location.
There are a variety of ways to derive a prediction for each location; each
method is referred to as a model. With each model, there are different
assumptions made of the data, and certain models are more applicable for
specific data—for example, one model may account for local variation better
than another. Each model produces predictions using different calculations.
The interpolation tools are generally divided into deterministic and
geostatistical methods.
The geostatistical methods are based on statistical models that include
autocorrelation (the statistical relationship among the measured points).
Because of this, geostatistical techniques not only have the capability of
producing a prediction surface but also provide some measure of the
certainty or accuracy of the predictions.
Kriging is a geostatistical method of interpolation.
b*e
16 楼
In GIS (geographic information science), spatial analysis to find best
location usually needs more than one data layer. For example, to find gold
mine in the mining industry, the GIS Specialist/scientist will consider the
geology data, hydrology data, soil data, vegetation data, elevation data (
slope, aspect, etc.), besides the existing location of gold mines.
location usually needs more than one data layer. For example, to find gold
mine in the mining industry, the GIS Specialist/scientist will consider the
geology data, hydrology data, soil data, vegetation data, elevation data (
slope, aspect, etc.), besides the existing location of gold mines.
T*u
17 楼
我的建议是尽量尽量跟多客户谈，听他们的意见，尽量用他们的经验，这种情况下没有
domain knowledge做model木出路。
domain knowledge做model木出路。
相关阅读
6/17开始的免费数据分析基础课程讲座data scientist opening (San Francisco)analyzing the analyzers最近有没有DS有关的会议啊？求Hadoop项目练手Rise of (Azure) Machine Learning这里有在用R做support vector machine的吗，哪个package用的比较多啊？Colah 关于 neural network 的一篇博客有人用过IBM的Predictive analytics产品么？这个烙印很牛逼么？想咨询一下洛杉矶这边data scientist起薪大概多少问一个 feature 相关性问题请教一道面试题~~data science 面试求教与其刷题，不如看美剧 (转载)大数据到底是不是忽悠？ (转载)跪求Amazon Research scientist onsite面经time sliced classification modelspython里面有没有generic的计算月底日期的函数？data science的personal statement怎么写？