4,934 reputation
11034
bio website www54.sap.com/industries/…
location Switzerland
age 38
visits member for 3 years, 7 months
seen 2 hours ago

During the day, I forecast sales at supermarkets, drugstores, furniture, perfume and other retailers and calculate order proposals. Lots of time series, with an emphasis on fast, automatic and robust data cleansing and forecasting - with some logistical optimization thrown in for good measure. I'm active in the International Institute of Forecasters and an Associate Editor for their practitioner-oriented journal Foresight.

At night, I switch hats and do inferential statistics for clinical and biological psychology.

I'm never bored. And I use R.


Apr
11
comment What are differences between the terms “time series analysis” and “longitudinal data analysis”
You do have a point re "causal" vs. "association", and of course longitudinal data can be used to forecast - it's just that I don't often see the two concepts together. Forecasters usually talk about time series. Apart from that, I couldn't put it any better than @gung.
Apr
11
comment What are differences between the terms “time series analysis” and “longitudinal data analysis”
This may turn into a poll... I have worked on both types of data, and one key difference seems to be that longitudinal data is often used in causal analyses, to understand the impact of interventions or treatments, whereas time series are often used in forecasting. Of course, the difference is not clear-cut (you need to understand the underlying drivers to forecast, and IMO you haven't understood the drivers unless you can forecast well). But people who do signal detection in time series often don't care so much about forecasting, so they would probably reject my distinction.
Apr
11
revised T-test homework
added self-study tag, removed mathematical-statistics tag
Apr
11
comment T-test homework
This looks like homework or self-study. We will happily help you, but we would like to know what you have already tried yourself and where you got stuck.
Apr
11
comment How do you extract confidence intervals and OR out of the step() function in R?
Confidence intervals make no sense any more after stepwise variable selection procedures (nor do p-values), unless you calculate them based on bootstrapping the entire procedure, including the stepping. There is an extensive literature on this, which I heartily recommend that you delve into. ORs may make sense, on the other hand (but you will be too sure of them after stepwise model building).
Apr
11
comment Data mining techniques in R for advertising and sales data
Thx. I adapted my answer to your data.
Apr
11
revised Data mining techniques in R for advertising and sales data
improved after OP provided data
Apr
11
revised Data mining techniques in R for advertising and sales data
Added time-series tag
Apr
11
answered Data mining techniques in R for advertising and sales data
Apr
11
revised Data mining techniques in R for advertising and sales data
corrected reference to HoltWinters
Apr
10
comment How to further reduce predictive error in a regression tree model
Have you looked at random forests? R offers the randomForest package.
Apr
9
comment Developing an appropriate time series model to predict sales based on past month record
@RobHyndman: did you want to answer Aksakal with your last comment?
Apr
8
reviewed Approve suggested edit on Question about the error term in a simple linear regression
Apr
8
reviewed Close Given two independent normal random variables $X$ and $Y$, what is $P(X\leq x\mid X>Y)$?
Apr
7
comment bootstrapping for a parameter estimation
As I wrote above, I don't understand why the minimum of (a truncated subset of) your sample should carry any information about the expected value, which you wish to do inference about, bootstrapping or not. So I don't see how your approach would be valid statistically in the context of making inferences about the expectation. But perhaps I'm just dense.
Apr
7
comment bootstrapping for a parameter estimation
I'll wait for the statistics gurus to chime in, but my suspicion is that the answer to your first question is No, as in here: en.wikipedia.org/wiki/Not_even_wrong Sorry. I'm open to being convinced otherwise.
Apr
7
comment bootstrapping for a parameter estimation
I'm not entirely clear on what you are asking. Why should the minimum of your sample carry much information about the expected value? Why do you cherry-pick data <0.9?
Apr
7
comment Standard mean and deviation
What have you already tried?
Apr
7
revised Standard mean and deviation
added self-study tag
Apr
7
comment Predicted values for fixed effect quantile regression
Please edit your question to contain a reproducible example. See here: stackoverflow.com/questions/5963269/…