89,793 reputation
8150312
bio website quantdec.com
location Northeastern US
age 14
visits member for 4 years, 2 months
seen 2 hours ago

Consultant (environmental and spatial stats a specialty), expert witness, and teacher. I can be reached through (outdated but still valid) links posted on my web site.

Twitter: @WilliamAHuber // ASA-P website: http://amstatphilly.org/


Why waste time learning, when ignorance is instantaneous?

--T(iger) Hobbes.

For any complex problem there is a simple solution. And it's always wrong.

--[Mis?]attributed to H.L. Mencken by Dava Sobel, Longitude.


2h
comment Shape uncertainty of a 3D point cloud
Please explain what you mean by "shape uncertainty." Some indication of how this point cloud was obtained and measured would be helpful.
2h
comment Two Gaussian Likelihoods with Two Decision Boundaries , 0/1 loss function
An "example" of what? What exactly would constitute an answer to your question?
2h
comment Best statistical method to show different performance week over week
+1 Thank you for this focused, thoughtful response.
2h
comment The Mean of the number of times until an event occurs
A closely related question is the Coupon Collector's Problem, in which blue cards are never replaced by red cards. Many of the techniques used to solve that problem apply to this one.
17h
comment How to prove that the Fourier Transform of white noise is flat?
What happened to $\gamma$ in the initial formula?
17h
comment Looking for and dealing with collinearity in a GLM
That sounds like it might be the question you should be asking, Hannah. How you treat your controls in the analysis is crucial and it looks like it might be causing the technical problems you are experiencing.
17h
comment Expected lifetime of a device with two parts each having spares?
Congratulations on reaching 10K, Dilip. (You did it with remarkably few answers, too.) I am glad to see this addition to our high-rep subcommunity, whose growth is essential to the health of this site.
17h
comment Is normality testing 'essentially useless'?
You might start with combining the two posts--the answer and your comment--and then think about weeding out (or relegating to an appendix or clarifying) any material that may be tangential. For instance, the reference to undefined means as yet has no clear bearing on the question and so it remains somewhat mysterious.
17h
comment Is normality testing 'essentially useless'?
Thank you. The remarks in that comment seem to be better focused on the question than your original answer is! You might consider updating your answer at some point to make your opinions and advice more apparent.
17h
comment Looking for and dealing with collinearity in a GLM
The "Control" column stands out and will cause problems: look at all those zero entries. What exactly does this represent? (It doesn't look like a proper control in the sense of a suitable reference for comparison, because there has been no replication.)
17h
comment Troubles reporting transformed variables for log and sqrt into a general equation
Non-linear re-expressions of the dependent variable will directly affect the variance of the residuals as well as the linearity of relationships with other variables. There are methods to identify appropriate transformations, based on an analysis of the original residuals. Transforming the independent variables has only an indirect effect and so there are few comparable techniques. I posted a case study leading to transforming both variables at stats.stackexchange.com/a/60455. As you can see, it takes a fairly special circumstance to warrant a re-expression of an independent variable.
17h
comment Troubles reporting transformed variables for log and sqrt into a general equation
Since there are many thousands of posts on our site about this topic, I recommend narrowing the search. A search on transformation regression dependent seems to focus the links pretty well. A very general, highly-voted thread at stats.stackexchange.com/questions/298 provides an overview in the context of a contemplated log transformation, but most of what it says applies to any nonlinear transformation. Another keyword worth searching on is Box-Cox.
18h
revised You will randomly select 10 balls from the box with replacement what is $E(\bar X)$
edited tags
18h
comment You will randomly select 10 balls from the box with replacement what is $E(\bar X)$
You're quite right: but it's not a trick, it's an important technique to know. Since it is obvious you have been asked this question in a textbook or course, then you must already have been exposed to some properties of expectation and variance that allow you to infer their values for sums of random variables: use those properties to solve this problem.
18h
reviewed Reviewed Aggregating Results for pvclust
18h
comment Aggregating Results for pvclust
If you are asking for the R commands to do this, then please flag your post for migration to Stack Overflow. If not please explain for non-R users what pvclust does and what kind of clustering it is.
18h
reviewed Reopen You will randomly select 10 balls from the box with replacement what is $E(\bar X)$
18h
comment Troubles reporting transformed variables for log and sqrt into a general equation
That helps clear things up, thank you. However, transforming an independent variable has no (direct) effect on the residuals. It will change the nature of the relationship (from a linear one to nonlinear, or vice versa), but usually is not done--and never is the first thing done--specifically to resolve problems with heteroscedasticity. Thus, although this does not address your question, I would like to ask whether you have first considered transforming $Y$ and, if not, to suggest that you try that before you go any further.
18h
comment How to prove that the Fourier Transform of white noise is flat?
Please edit your question if you wish to change it--not everybody will read these comments and most people will answer only what is in the question itself.
18h
comment What is the Fourier Transform of a brownian motion?
It's just integration by parts: see mathworld.wolfram.com/FourierTransform.html starting at equation (34).