meta user
network profile
| bio | website | www54.sap.com/industries/… |
|---|---|---|
| location | Switzerland | |
| age | 37 | |
| visits | member for | 2 years, 8 months |
| seen | 15 hours ago | |
| stats | profile views | 431 |
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.
|
May 2 |
comment |
Classification with 3 groups, repeated measurements, missing values, more predictors than subjects Thanks! I understand why a Rasch model would prefer data to be integers (it was developed to analyze questionnaire responses, after all). So if I decide to go this way, I will probably need to re-implement everything... |
|
Apr 18 |
reviewed | No Action Needed Is the number 20 magic? |
|
Apr 18 |
reviewed | Close Calibrating speed |
|
Apr 18 |
reviewed | Leave Open How to create an ADBUDG Economic Marketing Model with R |
|
Apr 18 |
reviewed | Leave Open Naive Bayes with Density estimation |
|
Apr 18 |
reviewed | Close Multi-armed bandit algorithms in Java? |
|
Apr 18 |
reviewed | Leave Open How to compare three different replacements for eggs in baking? |
|
Apr 18 |
reviewed | Leave Open Price elasticity of specialized goods |
|
Apr 18 |
comment |
How do I get better forecasts for this data +1. Especially for the suggestion about modeling the spikes. The data look a lot like daily retail sales with promotions (upward spikes) and availability problems, also known as "Out of Stocks" (the downward spikes). |
|
Apr 18 |
comment |
How do I get better forecasts for this data +1. However, given that there are only about two years of history, I wouldn't worry too much about leap years - any error introduced by using a 365 day period instead of a 365.25 days one is probably swamped by the noise in the series. |
|
Mar 6 |
comment |
Weighting the response variable in an lm This should be exactly what R does if you specify weights=1/slope.var. However, @AdamO points out a more appropriate method to deal with your question. |
|
Mar 5 |
comment |
Weighting the response variable in an lm Could you clarify how exactly you want to "weight your response variable"? The weights parameter in lm() yields parameter estimates that minimize the weighted sum of squared residuals ("normal" OLS minimizes the unweighted sum), so high weight observations will be more influential in estimating the parameters. This actually sounds like just what you should be doing in your problem, but I probably am missing something. |
|
Feb 21 |
reviewed | Leave Open Paired Samples T Test Suitability |
|
Feb 21 |
reviewed | Approve suggested edit on How to describe the differences in skewed data with same median but statistically different distribution? |
|
Feb 21 |
reviewed | Approve suggested edit on probability of one random variable being greater than another |
|
Feb 20 |
comment |
How fit a regression model with two time series @Glen_b: maybe I am misunderstanding your comment, or maybe you are misunderstanding my comment... I was not claiming your original comment was a subset of the answer already posted below - I wanted to suggest that you post your comment as an additional answer. Sorry for the confusion. |
|
Feb 20 |
comment |
How to simulate with given probability? What happens if the N(0,1) does not happen? A deterministic zero? |
|
Feb 20 |
comment |
How fit a regression model with two time series @Glen_b: isn't your comment already an answer? |
|
Feb 20 |
revised |
How fit a regression model with two time series retagged |
|
Feb 20 |
reviewed | Close How can I say this is probability? |
